Subscribe

Analysis EO
Analysis · 2
report
Analysis EO
Apr 19, 2020 07:00 am ·

Why Is There No AI Unicorn Like SenseTime in the US?

SenseTime, a Chinese AI startup established in 2014, has claimed (via their CEO Li Xu) to have a post-money valuation of USD 70 million, after an investment of USD 1 Billion from SoftBank. As an AI algorithm provider, it developed Computer Vision (CV) technology for multiple industries, including autonomous vehicles, security and fintech, after becoming famous for facial recognition when it was established. CB insights released a list of the most valuable AI unicorns around the world and evaluated SenseTime at USD 4.5 billion in 2019. There was only one US company – Tanium – that had a higher valuation (USD 6.5 billion). Unlike SenseTime, which acts as an AI algorithm provider serving customers from various industries, each US unicorn with valuation higher than USD 3 billion has a specific focus on one application area. So why is there no full-stack AI platform unicorn like SenseTime – a company that offers commercially applied AI algorithms for players in a wide range of emerging industries – in the US? Detailed analysis of SenseTime’s evaluation The SenseTime Story Since its founding as an AI facial recognition algorithm specialist, SenseTime has become known for its professionalism in AI technology research.  With a founding team from world famous universities including MIT, Tsinghua University, and CUHK, SenseTime has become one of the most prolific institutes regarding paper publishing in top AI conferences, including CVPR and ICCV. SenseTime also shaped its public image by various types of connection with Universities and industrial giants, including establishing AI labs, building academic alliances, and making technology commercialization collaborations. SenseTime is the prototype of startups that rode the AI wave in China since 2015. Propelled by Beijing AI-related industries have developed at a rapid speed, bringing a list of AI technology companies to public attention. SenseTime, along with similar companies like MEGVII and YITU, has been chased by hot money since then.  In September 2018, SenseTime was asked to build China’s next generation national AI open innovation platform. Analysts view it as the symbol of entrance into the Chinese ‘National Team’ along with tech giants Alibaba, Tencent, Baidu, and iFLYTEK. With ample funding and attention from the public and government, SenseTime then extended its service to more general CV fields, including large-scale video data understanding/mining and image processing /augmentation, for a wide range of industries like public security, autonomous vehicle and healthcare. SenseTime is facing more challenges as the commercialization of AI technology moves further. As it enters simultaneously into several industries, it inevitably competes with industrial giants, which have also developed related AI technologies over the years. In the public security industry, as SenseTime’s recent focused area, wherein AI technology is considered to have the most support from local governments, SenseTime is struggling to compete with Hikvision, the largest surveillance system provider in China and globally, which also has an applied AI technology accumulation (details for Hikvision and Chinese security industry). Picture of US AI unicorns AI startups, although also very popular in the US, have little chance to become unicorns without deep application of AI technology in a fixed industry. Comparing top AI startups in Silicon Valley and New York, the Bay Area AI startups generally have higher valuations because of their potential to apply cutting-edge technology to emerging industries. AI unicorns in the US generally have a clear focus on autonomous vehicles, biotech, enterprise services, etc. Startups like SenseTime that act as full-stack AI platforms for various industries are less likely to be recognized in the US because of the more developed and realistic venture capital market.  With a developed finance system, venture capital investments in the US are now distributed in a few areas with concentrated emerging industries, such as Silicon Valley. The more realistic investors require target companies to provide proven approaches to their ideas instead of arbitraging from hot concepts. So, AI startups betting on single industry and digging deeper are typically more favored. The Chinese venture capital market is, on the contrary, on the move, developing fast with ready investment money. Following the economy, cutting edge technologies with the potential to improve social efficiency are always treasured by the public and government. The undoubted benefit of AI technology to Chinese society, which is still not fully developed, is the real support backing SenseTime’s USD 3 billion historical funding. Dilemma of the Chinese AI unicorn The latest news has come out that SenseTime will not seek IPO opportunities in the near future on the Hong Kong Exchanges and is looking for PE financing instead (more details). This indicates that SenseTime is less confident in attracting global secondary market investors compared with venture capitalists. Meanwhile, Chinese venture capitalists in the AI sector are also getting realistic after witnessing the volatile investment market after 2018. The investors of SenseTime are mostly top long-term players with ample funding that aim to gain from the appreciation of SenseTime’s market value following the twisted but upgoing AI sector. It does not mean that investors are careless in SenseTime’s commercialization. SenseTime is now under the pressure of profitability, which can be seen from its recent focus in the Chinese government-led public security industry – a sector that is promising, but highly competitive. SenseTime seems to be in such a dilemma. On the one hand, commercialization requires it to first be rooted in one profitable industry, and government-led industries are the best choice in Chinese market due to Beijing’s firm support for AI; on the other hand, the public image of the Chinese AI algorithm unicorn stops it from fully turning to these industries as it is considered not transparent and not ‘high-tech’ by the mainstream capital market. SenseTime’s effort to enter most other industries is paying back slowly. The ToC market is monopolized by Internet giants like Alibaba, Baidu and Bytedance, as they have massive customer data and also the ambition to develop their own AI technology. Most ToB markets in China like automobile, smartphone and finance are also hard to penetrate, since the industrial oligarchs started to bet on similar technology even before the recent AI wave. There is still a long way to go before SenseTime finally becomes a full-stack AI platform that has deep cooperation with players in various industries. As an AI research startup, SenseTime is inevitably facing barriers when entering new markets. Maybe learning from its US peers and labelling itself as a specialist in a smaller industrial sector is a better choice, at least in the near future.

Analysis EO
Analysis · 2
report
Analysis EO
Apr 5, 2020 09:00 am ·

SenseTime to Stay Private Involuntarily

Source – Nikkei Asian Review (March 18). SenseTime (商汤科技) suspended its IPO plan on the Hong Kong Exchange. Instead, it would seek a new USD 1 billion financing series. Established in 2014, SenseTime, or say VCs, has made itself the most valuable AI company in the world valuing at USD 7.5 billion. SenseTime has raised nine rounds of financing, making a total of USD 3 billion. The latest financing series created a record in the industry, USD 1 billion for a single round. Many investors are big names in corporate funds and VC funds such as Qualcomm, Alibaba, Softbank, Tiger Global, IDG Capital, etc. It has witnessed the maniac AI investment time and now the No.1 AI unicorn is questioned for its over-inflated AI valuation. After six years of operation, SenseTime has been rumored for times of going public. While it remains private, another AI unicorn, Megvii (旷视科技) launched its IPO in 2019 with hurdles and tumbles: after passing the second hearing in Hong Kong Exchange, Megvii deferred ringing the opening bell due to the COVID-19 outbreak, which may cause its IPO application to expire. So far, China’s ‘Four CV (computer vision) Dragons’ – SenseTime, Yitu (依图科技), Megvii and CloudWalk (云从科技) – are staying private. Apart from Megvii, each of their IPO timetables is yet to be clarified. Investment frenzy in AI sector ends The year of 2018 marked the craziest AI investment period. Many dollars flushed into the sector and lifted up AI startups’ valuations – AI unicorn companies appeared one after another with hundred-million level VC investments. Among them, SenseTime became the ‘shining pearl’ after it was given the highest valuation, and also the highest deal value recorded, the USD 1 billion Series D funding from SoftBank China. Not only in the AI sector, China’s VC investment heat across the board has been cooling after the peak in 2018. VC development requires a financial market that can provide as many exits as possible and a healthy and stable environment for startups to grow and scale. In 2019 we witnessed exacerbating global trade tensions between China and other countries: we saw certain of China’s companies pause infrastructure projects, watched fluctuating (most of the time increasing) tariffs on both sides between China and the United States, the creation of the US commerce blacklist, and so on. Chaos is stirring in the background, and some global investors like Tiger Global, are shifting their focus to other regions like India and Southeast Asia. Singapore’s AI companies received general investments from VC investors last year, and among these companies, Advance.AI was funded the most, landing USD 80 million in its Series C financing. Still, China is one of the biggest markets for VC investors, but it is not the only one. China has the most unicorn companies in the world – 206 unicorns as of 2019 – while the United States has the second most unicorn companies, 203. What is SenseTime doing? Backed up by billions of dollars, SenseTime founded its own corporate fund to invest in AI startups and in the meantime, according to Tianyancha’s record, it has founded tens of AI companies with support from regional governments. SenseTime has cooperated with provincial and civil education departments to advance AI + Education mode. As early as late 2018, the Qingdao government signed an MOU with SenseTime to promote AI education, and starting from September 2019, Qingdao launched the AI courses in K12 education with over 20,000 students involved. As reported by Reuters, SenseTime expected USD 750 million for its 2019 revenue and we have sufficient reason to believe that major clients of SenseTime are from the government side. Excepting education, SenseTime’s revenue mainly depends on security-related products and services. Under the rising concept of the ‘Smart City,’ SenseTime has cooperated with nearly 100 cities in China to support their upgrading projects. However, the ticket price to be a Smart City partner is never zero. In 2017, SenseTime signed an MOU with the Shanghai government, mentioning that the AI unicorn would invest no less than CNY 5 billion (USD 700 million) within five years, whereas expected revenue for 2019 was USD 750 million. The billion-level investment deal for a three-year-old startup has a different meaning: a deal for a deal. SenseTime soon founded its Shanghai branches and pushed AI-related projects with support from the top. We can expect similar cases when it cooperates with some of the other regional governments. How many deals or contracts can a CNY 5 billion investment bring back to SenseTime? How about several CNY 5 billion investments? By looking at SenseTime’s current performance, its investors are not able to profit from revenue earnings – but surely can be well paid by successor investors as the valuation is pumped. Slimming and deliberate trying Indeed, security and Smart City are critical for SenseTime at this stage and they contribute significantly to the revenue. The company has been seeking non-government side opportunities. Smartphone OEMs are one of the targets. AI companies provide facial recognition solutions for smartphones and the business seems to be promising: yearly global smartphone shipment exceeded 1 billion units for the past five years. The cooperation was a good deal until smartphone OEMs began setting up their own teams to develop facial recognition solutions and other computer vision functions. Hardware manufacturers with AI teams will be a headwind for ‘soft’ AI companies on their way to commercialization. Along with the fierce competition, public concerns about data security also shadow the way ahead. SenseNet (深网视界) was a company invested in by SenseTime. A data leakage scandal occurred in 2019 and destroyed the company’s reputation. SenseTime swiftly got rid of the company. Besides SenseNet, SenseTime also sold all its shares in Tangli Technologies (汤力科技) and reduced investments from co-founded companies to get slimmer. Starting from CV, SenseTime has developed its business in many areas, including but not limited to fintech, mobility, healthcare, education and advertising. At the early stage, we can see a clear pattern in China’s ‘Four CV Dragons:’  SenseTime focused on public security, Yitu was famed for its efforts in healthcare, Megvii worked on facial recognition and CloudWalk pioneered in fintech. Now, they have all expanded to multiple areas and their product mix is transformed from vertical focus to fuller coverage. Expansions are on the way. For SenseTime, 2019 is a year for attempts in education and smart cities. The most valuable AI unicorn does not have a clear timetable for its IPO. As Megvii has been tumbling on the winding IPO track since mid-2019, SenseTime has remained private. The overinflated valuation problem makes all AI companies into embarrassments before they can truly stand on their own feet. In this sector, companies’ profitability issues concern most public investors, who care less about passionate stories but emphasize more real financial statistics. When the current ‘Black Swan’, COVID-19, hit all industries worldwide, investors became even more cautious and conservative. Raising money in either the private market or public will be more difficult than before. Many of Masayoshi Son's portfolio companies such as WeWork, OneWeb and others are in turmoil –  returns below expectations or holding on their last breath. SoftBank's used-to-be-shining tech startup portfolio is dimmer than ever. Meanwhile, SoftBank-invested SenseTime’s IPO will not be anytime soon given an environment of declining market confidence and a global economic recession – SoftBank’s ‘curse’ arising from its portfolio companies’ IPO tracks continues.

Analysis
Analysis · 1
Analysis
Analysis · 1
Analysis
Analysis · 1
Analysis EO
Analysis · 2
report
Analysis EO
Jul 29, 2019 07:00 pm ·

Beyond the STARs: Meet Overhyped Young China's Chipmakers

The SSE (Shanghai Stock Exchange) Sci-Tech Innovation Board, known as Star Market, finally opened for business this Monday, facing an insane demand jacking up the share prices that initially were quite high per se. At the end of the first trading week, all the 25 companies got their stocks bloated. The current situation is apparently unhealthy and it is almost certain that the real value is way below the figures we are observing. The market is chaotic and might crack any time, triggered by the countless amateur investors that behave irrationally.  "Let the Star board be a bubble, then let it burst. That could be a first step in the way to a more mature financial market in China." Wayne Shiong, partner at China Growth Capital Although individuals with less than two years of trading experience are not allowed to invest on the Star board directly, it doesn't seem to be an insurmountable barrier for them: there is always a friend of a friend that may help. Apart from that, the number of low-qualified but "experienced" investors is historically high in China, a country of almost 1.4 billion individuals with a frail social welfare system. The local rule is simple: retirement is one's own issue. Chips in the stack Despite the inherent volatility of mainland China markets and sky-high valuation of the first batch of tech firms that went public on the Star board, there is a category of companies that, due to a plethora of factors, look flexible – semiconductors. All the four stocks representing the industry could be found among the sci-tech board's top performers now.  AMEC (Advanced Micro-Fabrication Equipment Inc China), Anji (Anji Microelectronics Technology Shanghai Co Ltd), Espressif (Espressif Systems Shanghai Co Ltd) and Montage (Montage Technology Group Ltd) had been in good standing well before the new platform started trading on July 22. All their offering P/E ratios were above 40.02 – market average on ChiNext, another China's Nasdaq-style board, let alone AMEC's (688012:SH) record-high 170.75. Wild market ingested chunks of stocks put up for sale by the primary investors. According to the Shanghai Stock Exchange, more than 50% of organizations that obtained share allocations during the 25 public offerings happened earlier this month sold their stakes on the first day of trading.  Three open trading sessions turned to be enough for the initial craze to cool down and come to the delicate equilibrium. On Tuesday, volatility dropped; it didn't fade away, though. According to EqualOcean's evaluation, the second phase (post-psi) of the new venue's evolution process started in the afternoon trading session, July 23. From this time on, market players began to be less spontaneous in their decisions. Anji Technology (688019:SH) remained to be an object of market speculation in the post-psi phase; however, it became almost three times less volatile. Espressif Systems (688018:SH), a WiFi chipset producer, also slightly reduced the density of fluctuations, staying vulnerable to dodgy public decisions. In general, semiconductor companies' stocks have been growing actively throughout this week. For a small (255 employees as of 2018) Montage Technology (688008:SH), which collected a revenue of USD 1 million per employee last year, this boost brings a reason to be rightly called a mid-sized chipmaker. Its recent stellar financial performance (for one, net profit margin of 41.2% was recorded in 2018) positively affected the stock price. The other three firms have also been reporting profits since 2016.  Worth noting that all the "Star chipmakers" are Shanghai-based, their combined market capitalization climbed up 186% from CNY 50.63 billion (USD 7.36 billion) to CNY 144.86 billion (USD 21.06 billion) within last week. Although this isn’t an exception as all the stocks on the Star board have surged remarkably, there are some reasons why chipmakers gain more attention than other outperformers. "Current lofty valuation of semiconductor stocks on the Star Market is reasonable: investors of all types are positive about this industry in China." Wang Liming, executive director of CICC The four firms mentioned all are strong industry players having distinctive competitive advantages at certain stages of various value chains. Nonetheless, their "chipmaking identity" is a critical factor in the market assessment. This statement was approved last week as the SSE sci-tech board started trading. EqualOcean undertook a systematic approach in dissecting the situation around the design and fabrication of semiconductor devices and microelectronics businesses. What's behind? According to McKinsey estimation, semiconductor companies so far obtained less than 30% and 20% of the value coming with technology stack in PC and mobile areas respectively. At the same time, software developers grabbed a more significant hunk. The consultancy bets that this proportion will flip over in the near future, facing chipmakers taking home up to 50% of the technology stack-related cash. Why? Most of the major forces swaying the semiconductor sector in China nowadays can be broken up into five groups by a scale in time: "global zeitgeist", "local zeitgeist", "state strategy and policy", "geopolitical tensions" and "industrial landscape". Some of the categories mentioned are unwrapped below. Connectivity, a truly monstrous trend, has become a major game-changing wave globally at various levels from household IoT (Internet of Things) to global communication and transportation. Last year, Bain&Company estimated that worldwide IoT market size will hit USD 520 billion by 2021. The so-called "digital transformation" further accelerates the ubiquitous IoT adaptation process.  Tech improvement and sustainability needs are paving the road to more environment-friendly methods of transportation. Besides, car makers have lately been involved in newly emerged battlefields of the non-price competition. Striving to enhance road traffic safety -- the realm seeing most vicious market rivalry today, -- they spur the rapid growth in global sales of automotive semiconductors.  Besides global zeitgeist Surely, planetwide trends do shape investors' expectations. There are some issues that are special about China in this regard though. IIot (Industrial Internet of Things) is hot this year not only in tech media but in business circles as well. As the overall quality of goods produced by the "World's Factory" improves, local companies look for smarter ways to organize the manufacturing process. Even at the micro-level.  Why China is a promising market for household appliances is beyond question. And IoT goes hard therein recently. Local tech giants like long-suffering Huawei and "young and passionate" Xiaomi (01810.HK) are actively restructuring the product mix in order to reap the fruits from the country's growing middle class.  The world's largest passenger car market is where fierce competition between producers in the automotive industry happens. In recent years, we have experienced a boom of EV (electric vehicles) sector that was sparked by various objective and subjective factors (read here about how and why EV industry was sprung up in China). Autonomous driving also contributes to the burgeoning but brutal auto scene. And this is where microelectronics is pivotal. Public security has been a viral issue in China for a long time. The government maintains close ties with local private firms that position themselves as "AI companies". For example, closely followed by EqualOcean algorithm-developer SenseTime is among them. The public-private partnership segment in this industry got saturated in 2017, but a spacious room for deepening relations exists internationally (say, the famous Equador case). No doubt that China, with its relatively nimble, progressive state-management model, bets big on the cutting-edge technology. Microdevices will become a core element in the country's near future, and chips are essential in any of those.  "The Chinese government has been supporting domestic semiconductor firms extensively and will continue doing that -- this is a long-term trend." Bo Chunmin, senior analyst at EO Intelligence The next dimension to put on the table is the current geopolitical backdrop. Numerous uncertainties evolved recently as the world shifted from the organization-based to the agreement-based model. While the global division of labor is being staggered by irrational and spontaneous political decisions, countries have no choice but to prepare for the worst. China isn't willing to suddenly find itself in zugzwang; the government thereby makes changes accordingly. Supply chains become shorter and more localized. Ecosystems, being exposed to spillover effect and "liquid" innovation, expand rapidly. Accounting for up to 90% of the worldwide smartphone supply, China still imports core components.  Huge space exists in the industry, and a broad gamut of investors from state-backed organizations such as CICIIF (China Integrated Circuit Industry Investment Fund) to venture capitalists want to participate and frame the new technological revolution.  In short Successful IPO is a critical milestone for the nascent Chinese semiconductor companies. Their current valuation and low volatility rates validate the market's trust that is originated from three converging pillars: global connectivity trends, the industry's strategic status in the country and their splendid financial performance. EqualOcean considers chipmakers' stocks as the least fragile among those on the Shanghai Stock Exchange Star Market to date.

Analysis EO
Analysis · 2
report
Analysis EO
Jun 21, 2019 08:52 pm ·

The Unicorn SenseTime to Incubate the Next Gen of AI Startups

SenseTime, the idol artificial-intelligence (AI) tech unicorn in the world, grows from an AI enabler into an integrated AI institute in the industry. SenseTime is no longer simply a company with cutting-edge deep learning (DL) and AI technologies as presumed at its birth. Stepping into the fifth year, SenseTime scaled up to a company with a size of over 3,000 employees. With different verticals and functional departments, SenseTime strives to maximize its enabler identity and realizes AI into real practice through varied means. It is known that SenseTime raised rounds of funding and was valued USD 6 billion at the latest financing round. Being an investor, SenseTime is not as famous as its financial backers like SoftBank, Alibaba, Temasek, etc. Reported by Bloomberg, SenseTime set up a fund of USD 2 billion to invest AI-relate startups. It is considered practical and realistic being an investor. Shown on itjuzi.com, SenseTime has invested or acquired 11 companies dated now and most of these financing activities are strategic/early-stage investments. From SenseTime's investment trait, the direction of SenseTime’s development is revealed. Computer vision (CV) is the starting point of SenseTime and so its early investment cases. Linkface, Xinedge, and SenseNets were merged into SenseTime and they were highly CV-tech-related companies. Computer Vision Investment Era Linkface was a company specialized in facial recognition technology. With strengths in search engine optimization and self-developed facial nodal points recognition algorithms, its technology has been applied in numbers of financial institutes for identification, anti-fraud and other scenarios. The merge enhanced SenseTime's capability in facial recognition technology. Though with strong soft capabilities, the development of image input hardware restricts the upper bound of algorithms and applications. Xinedge is a company with the ability to design and produce image input devices based on the need for computer vision. The image input devices are cameras that can optimize the quality of the captured images and return a better result for the image analysis process. The two companies introduced above are CV tech companies, so did others that SenseTime invested or acquired before 2017 share the same characteristics - CV tech specialized. From monotone CV companies to a diversified portfolio, SenseTime has broadened its horizon and eyed on expanded potential portfolio range since 2017. Evolution of Investment Suning Sports and Moviebook are the two seemed less related to SenseTime’s core business at the time. Suning, originally a physical electronics retail chain in China, has evolved into an integrated group. Suning and SenseTime’s cooperation is intense and mutually benefited since they are each other’s investor. Along with the trend of intelligentization, Suning is a pioneer in the Smart Retail transformation wave, which cannot be separated from the tech assistance from SenseTime. Suning Sport was invested by SenseTime in July 2018, which was only three months after Suning invested in SenseTime’s series C financing along with Alibaba, Temasek, and Beyond Ventures. Suning Sports’ main operation is sports game broadcasting. How SenseTime can be related to broadcasting cannot split the video visual technology, which is the strength of Moviebook. Moviebook enables the commercial service by means of AI imaging analysis and embedding advertisement into videos through its visual technology. Zhiruyi (植入易), it means easy commercial insert) helps to insert either still or moving commercials into videos; Videoyi simplifies the video analysis process such us element tagging, identification, tracking, etc. Moviebook represents the future of AI+ entertainment function and Suning Sports is a representative that could be re-formatted entertainment subject. Diversification As late as in 2018, SenseTime’s portfolio map demonstrated possible AI+ industries at the time. From CV to healthcare, the AI giant has been trying to explore as many applicable scenarios as possible. Terminus was previously a smart lock designer and producer and it has upgraded to an inclusive security service provider using its smart building technologies. Dated now, Terminus is a unicorn-sized company valued at USD 1.1 billion from its latest financing closed in 2018. Building an IoT-enhanced security ecosystem and AI technology to facilitate the soft part of the ecosystem is the strength of Terminus. Terminus AIoT Dashboard. Source: Terminus The unicorn, SenseTime, investing in smaller unicorn like Terminus indicates a mindset transition. At an emerging market, top players will strive to dominate the market and optimize their impacts. For smaller players in the fast-growing market (i.e. AI-related markets), they are less seen but still with strong expertise in some certain areas that can support them staying in the game. Investment in Application The top four AI tech companies in China, SenseTime, Yitu, CloudWalk, and Megvii, are expanding speedily. Yet from competing with peers, they focus on selves’ development and actively explore the boundaries of the applicable scenarios. For instance, SenseTime is well-known for is security ecosystem and Yitu is famed for its healthcare products, but their AI solutions have covered from smart city to smart manufacturing. As we can tell from SenseTime’s investment trait, the border of its investment is expanding along with the growing business. Based on the observation of SenseTime’s investment activities, most cases are the applicable side of AI technology. Truly, the commercialization and sustainable business development are the two most-cared topics for AI unicorns and interest involvers. It is evident to conclude that SenseTime’s investment activity is at phase I or at least the application phase since it focuses on AI tech application scenarios and these-and-such applications are critical to its short-term sustainability and the completeness of business model. For the next investment phase, the strategy might evolve to a higher level, such as investing in computation hardware and others that can provide a firmer foundation and wider playground for SenseTime to grow and upgrade. SenseTime breeds the next generation of AI startups via strategic investments and is building an inclusive AI eco-system through partnerships with these newborns and baby unicorns. Similarly, another top player, Megvii (previously Face++), invested several startups such as Video++ (极链科技), WhaleHouse (鲸仓科技), XianLife (鲜生活), etc. in last year. AI tech unicorns investing in startups that are closely related to AI tech applications is common in China. Investing in these startups is also a way to invest themselves.

Analysis EO
Analysis · 2
report
Analysis EO
Jun 17, 2019 10:36 pm ·

Introducing Top-10 Chinese AI Frontline Players

After Geoffrey Hinton proposed the concept of deep learning in 2006, a new epoch of artificial intelligence (AI) development began. Thousands of AI startups have emerged in the past decade. According to Wuzhen Institute, a Chinese non-profit research organization, the amount of AI enterprises in the world had reached 15,916 by the end of 2018. The U.S., China and U.K. topped the world with 4,567, 3,341 and 868 AI companies respectively. AI has become a target area for global investors. Venture funds, internet giants (such as BAT) and other big-scale actors joined the strategic-investment race in order to benefit from AI in the future. According to ITjuzi enterprise database, that the amount of AI-related investment events in China hit 643 in 2018, accounting for 9.7% of the total number of financing rounds over the same period in the country. Four years ago, only 158 events of this type occurred - the AI share in the whole funding pie was almost three times smaller. This year, we expect it to surpass 10%. AI-related investment events occurred in China making up 3.3% in total. This number is expected to surpass 10% this year. Dozens of AI unicorns have appeared one after the other in China, commonly with several rounds of financing and raising billions of money in total. This article summarizes the top 10 AI enterprises by the number of funding rounds. SenseTime Founded in 2014, SenseTime has grown to be one of the most successful AI unicorns in China and gained a good reputation in the world. Its valuation has skyrocketed to CNY 39 billion (USD 5.63 billion) after 10 rounds of financing by well-known investors from around the world, Tiger Global, Silver Lake, Fidelity, SoftBank, Qualcomm, Alibaba, Temasek Holdings and so on. The core competitive advantage of SenseTime is advanced deep learning technology, especially in computer vision. Its best-in-kind R&D team consists of Ph.D. holders and college professors from MIT, Chinese University of Hong Kong, Tsinghua University and Peking University as well as tech experts from Microsoft, Google, Lenovo and Baidu. Both of two founders, Tang Xiaoou (汤晓鸥) and Xu Li (徐立), are deep learning scientists from The Chinese University of Hong Kong. With the strategy of “AI for everything”, SenseTime develops AI solutions for multiple industries and areas, including smart city, smartphone, internet entertainment, ads, auto, finance, retail, education, real estate and so on. The company has cooperated with more than 700 well-known global institutions and enterprises, such as MIT, Qualcomm, Nvidia, Honda, SNOW, Alibaba, Suning, China Mobile, OPPO, Vivo, Xiaomi, Weibo, Wanke, etc. Though the specific revenue or net income figures are not disclosed, researchers from EO Intelligence - an industry and technology research institute - confirmed that in 2017, SenseTime posted revenue of around CNY 700 million and became profitable. The primary source of revenue is the sales of surveillance cameras with facial recognition. In 2018, sales reached CNY 7 billion, well exceeding all competitors'. SenseTime set its headquarter in Beijing and other offices in Hong Kong, Shenzhen, Shanghai, Chengdu, Hangzhou, Kyoto, Tokyo and Singapore. Ant Financial Ant Financial is one of the highest valued fintech companies in the world, officially founded in October 2014. After six rounds of financing, its valuation has surpassed USD 160 billion. The investors consist of many global financial behemoths and famous investment companies, such as Government of Singapore Investment Corp, Baillie Gifford, General Atlantic, Temasek Holdings, Sequoia, CPIC, Alibaba and so on. As an affiliate of Alibaba, Ant Financial was born from Alipay, a third-party mobile and online payment platform launched in 2004. The company is more like a technology enterprise with its strategy called BASIC (Blockchain, AI, Security, IoT and Cloud computing), raised in 2017 in a conference organized by Ant Financial.  Ant Financial focuses on payment, wealth management, P2P and technology service, with dozens of products such as Alipay, Yu’E Bao, MYbank, Ant Credit Pay, Zhima Credit, Huabei, Jiebei and so on. Till now, Ant Financial has acquired 14 financial licenses including banking, insurance, payment, securities, P2P and so on. According to Ant Financial, it will focus more on technology services as a third party. It is expected that 65% of its revenue will come from technology service in 2021. EO Intelligence predicts that Ant Financial will achieve a revenue ranging from CNY 94.2 billion to CNY 125.6 billion in 2021. ByteDance ByteDance is one of China’s internet giants operating several famous content platforms like such as Toutiao, Douyin (the domestic version of TikTok), Huoshan short video, Xigua short video and so on. Eight rounds of financing drive ByteDance’s valuation to CNY 487.5 billion, with famous investors like Sequoia, Softbank, DST, KKR, General Atlantic, Shunwei Capital and so on. The company was founded by Zhang Yiming (张一鸣) in 2012, headquartered in Beijing. Based on a precise recommendation system driven by machine learning, Toutiao and Douyin have attracted massive internet traffic. ByteDance disclosed that Toutiao’s monthly active users (MAUs) reached 254 million in last September, and the MAUs of Douyin surpassed 500 million by the end of this January. Beginning its global business from 2015, ByteDance has unveiled its products in over 150 countries with top ranking in more than 40 countries’ Apps markets. According to Zhang Yiming, the total MAUs of ByteDance’s products have exceeded 10 million by the end of March. Douyin jumped to the first in App market with 45.8 million downloads in the first quarter, followed by Facebook, Instagram and Youtube, reported by Sensor Tower. The main contributor to ByteDance’s revenue is ads because of its great number of users. According to an insider, the revenue of ByteDance has leaped from CNY 15 billion in 2017 to CNY 50 billion in 2018. Unisound Unisound is one of China’s leading speech recognition technology enterprises, founded in 2012 in Beijing. The company has closed eight rounds of funding, leading its valuation to USD 1.9 billion. The investors consist of several famous companies, such as Qualcomm, Qiming Venture Partners, CICC, JDD and so on. Unisound stands out on speech recognition, speech synthesis, knowledge graph and other speech-related technologies. As an AI solution provider like SenseTime, Unisound provides customized services for different clients from various industries. In the field of smart home, Unisound provides speech operation systems to several home appliance giants, including GREE, Midea, Changhong, VATII, etc. In healthcare, Unisound debuts text-to-speech solution UniHealth to doctors, a more efficient way to input medical records merely by speaking other than typewriting, employed in about 100 hospitals. Unisound also provides its technology to vehicle interaction device producers and education providers. To improve the performance of its products as well as reduce costs, Unisound unveiled AI chip in last May. According to Huang Wei (黄伟), founder and CEO of Unisound, the company has provided services to over 20 thousand companies covering 200 million users. It is expected that Unisound’s revenue will surpass CNY 500 million by the end of this year, with profit of over CNY 100 million. UBTECH Founded in 2012, UBTECH is one of the global leading service robotics companies, based in Shenzhen, Los Angeles and Silicon Valley. The company raised seven rounds of financing with the newest valuation of USD 10 billion, behind which are investors including Tencent, IFLYTECH, BYD, Green Pine Capital Partners, Minsheng Securities, etc. Servo steering engine, the motion controller of the robot, is the cutting-edge technology of UBTECH. The company designs and produces the advanced servo steering engine. Based on it, UBTECH develops humanoid robots with better motion ability compared with other similar products. The company’s R&D team has over 500 experts in robots and artificial intelligence, holding 1,253 patents by the end of 2018, in the field of Servo steering engine, computer vision, natural language process, SLAM, and so on. UBTECH has developed various types of robot products for entertainment, coding education, public services and security, with multiple use case scenarios, including home, retail store, train station, airport, bank, hospital, school, hotel, etc. For instance, its public service robot Cruzr has been employed in more than 2,000 Easyhome stores. This event is regarded as the biggest commercial case for service robots in the world. According to Wang Rulin (王汝林), COO of UBTECH, half of the revenue comes from foreign countries, especially in North America and Europe by collaborating with Best Buy, Apple, Amazon, Media Markt and other sale channels. The company is preparing to list on STAR Market this year. Megvii Megvii, led by three computer scientists from Tsinghua University, is one of the global leading AI giants providing AI solutions for various industries like SenseTime. With seven rounds of financing, this company’s valuation has reached USD 4 billion invested by some well-known global companies and investment institutions, such as Macquarie Group, ADIA, RCFSI, Foxconn, Alibaba, Sinovation Ventures, etc. Founded in 2011, Megvii started its business with Face++, a facial recognition solution, mainly used for identification in internet finance companies. For instance, Alipay is one of its important clients. Based on its facial recognition technology, Megvii develops smart surveillance cameras for government clients, which contributed a lot to the company’s revenue before this year. However, after the intense competition among several smart camera suppliers, this market has limited room for further development. Thus, the company launched its new strategy of IoT last year, providing IoT solutions for home appliances, city infrastructures and supply chains. In addition, Megvii also provides 3D facial recognition technology for smartphone, cooperating with eight mobile phone manufacturers including Huawei, Vivo, Xiaomi, smartisan and so on. The company is developing AR technology, whose main application scenario is the smartphone as well. According to Xie Peng (谢鹏), vice president of Megvii, the company has grown to be a over-2,000-employees AI lead player serving more than 1,000 companies in over 200 countries and regions, with some leading industrial leaders like Ant Financial, Foxconn, Lenovo, CapitaLand, CR Holdings, CHINA CITIC BANK and so on. Megvii planned to list on board this year – possibly for Nasdaq according to someone familiar with the matter but reconsidered the plan because of the trade war between U.S. and China. Video++ Video++, founded in 2014, is a Chinese technology company that creates more possibilities in the field of video advertisement by artificial intelligence. With 7 rounds of financing, the company has grown to be a unicorn, whose main investors include Alibaba Group, YF Capital, UBTECH, Megvii, Sirius Capital, Winshare Capital and so on. FaceAI, one of the core products of Video++, can understand the content of video and place advertisements in proper positions in videos automatically and in real-time. With Object recognition ability, FaceAI can understand 8 kinds of information in videos, including human face, place, object, brand, landmark, movement, expression and sound. According to Video++, FaceAI has served 873 kinds of scenarios, with more than 12 million minutes videos, including more than 120 top brands, such as imgo.tv, iQIYI, DOUYU, letv and so on. DONG Huizhi (董慧智), the co-founder and COO of Video++, said the valuation of this company has reached CNY 6.5 billion by the end of 2018, and the revenue grows rapidly in the past three years. Its annual revenue is only around CNY 3 million in 2016. This number changed to more than CNY 100 million in 2017, which mainly came from the service fee of advertisement. In 2018, Video++ created around CNY 600 million revenue under the overall arrangement in e-commerce. Yitu Technology Founded in 2012, Yitu Technology is one of China’s AI unicorns with seven rounds of financing. Its investors include reputable investment companies like Sequoia China, Banyan Capital, Hillhouse Capital, Zhen Fund and YF Capital. The company is known by the public mainly because of its face recognition system. Till now, Shanghai polices have arrested more than 60 suspects at the concert of Jacky Cheung with the help of the face recognition system provided by Yitu. In addition, the company has AI abilities including computer vision, speech recognition, natural language processing, as well as AI chips design. Yitu Technology focuses on three industries: security, finance and healthcare. In 2013, the company started to develop smart cameras with face recognition. Now, its smart cameras have been equipped in over 30 provinces in China, and Yitu has become the main supplier of smart cameras for the Shanghai government. In 2015, Yitu applied face recognition to payment and cooperated with several China’s banks like Agricultural Bank of China, China Merchants Bank, Shanghai Pudong Development Bank and Pingan Bank. In 2016, Yitu founded its subsidiary, Yitu Medical, to develop AI-assisted imaging diagnosis systems. Its core product, Care.Ai TM, is for a variety of diseases, including lung cancer, bone age detection, pediatric intelligent diagnosis, and medical record intelligent search engine. As of January 2019, more than 50 top hospitals in China have adopted this product. Horizon Robotics Horizon Robotics, a leading AI chips designer in China, has closed six rounds of financing from it was founded in 2015. Its existing investors include Hillhouse Capital, Linear Venture, CMBC Capital, Morningside Venture Capital and so on. Horizon's terminal computing focuses on two scenarios. One is for smart vehicles which will become a super terminal, and the other is for IoT devices, mainly cameras. Under smart retail scenarios, Horizon technology embedded-camera can substitute humans to monitor store situations, which can thus cut costs and improve efficiency for the retail industry. For now, bases of industrial division and cooperation have not been established yet, Horizon pursues to leverage others with its chips designed for the camera and smart vehicles as an underlying computing platform, according to YU Kai (余凯), the company's CEO and founder. YU also implied that the company's revenue in 2018 has reached more than CNY 100 million (USD 15 million) since the company released Gauss-architecture-based processor Sunrise 1.0 and Journey 1.0 in December 2017, growing 10 times compared to 2017. 2019 is the starting year of commercialization, according to EO Company. Cambricon Founded in 2016, Cambricon has grown to be a unicorn company in the field of AI chips, valuated USD 2.5 USD after its B series funding in June 2018. It has financed for 6 rounds, and its main investors include many state-owned large funds, such as CMCDI, China Venture Capital Fund, Chinese Academy of Sciences, etc. In 2016, Cambricon launched Cambricon-1A processor, which is the world's first commercial deep learning dedicated processor for smartphones, security surveillance, wearables, drones and smart driving. In 2018, Cambricon launched Cambricon 1M chip for edge computing, and the MLU100, the first in a new chip series for cloud computing. Cambricon will launch its second generation of AI chip, MLU270, for cloud data center servers this month, according to someone familiar with the matter. Last year, Cambricon launched the first-generation product, MLU100, and obtained purchase orders from Inspur, Lenovo, Sugon, DiDi and HIKVISION. Born from the Institute of Computing Technology Chinese Academy of Sciences, Cambricon has advantages on R&D of AI chips. CHEN Tianshi (陈天石), the founder and CEO of Cambricon, is an outstanding expert in the field of processor architecture and AI with a high reputation all over the world. He obtained received his Ph.D. in engineering from the School of Computer Science, China University of Science and Technology in 2010. He has attracted lots of scientists to join his firm, who have an average of more than 9-year R&D experience. Last year, CHEN Tianshi revealed that Cambricon might list in the A-share market in the future.

Analysis EO
Analysis · 2
report
Analysis EO
May 23, 2019 05:22 pm ·

Will SenseTime Get Behind the Wheel?

SenseTime, one of the domestic unicorns in the field of artificial intelligence, held the second AI summit, launching and showcasing a series of 11 products and solutions on May 15. In addition, it also introduces the layout of its automatic driving and intelligent vehicle system in detail. Founded in 2014 by TANG Xiaoou (汤晓鸥), professor of information engineering at the Chinese University of Hong Kong, the company specializes in the development of original technologies for computer vision and deep learning. As of September 10, 2018, SoftBank China has invested USD 1 billion in SenseTime technology co. LTD, raising its valuation to USD 6 billion. SenseTime has industrial and research facilities in Hong Kong, Beijing, Shenzhen, Shanghai, Chengdu, Kyoto and Tokyo. Strategic partners include Qualcomm, Huawei, Xiaomi, Honda and MIT. Computer vision and deep learning algorithms AI technology can support both autonomous driving and intelligent vehicle-mounted systems at some level. In addition to providing basic visual modules for intelligent network vehicles, SenseTime can also help manufacturers build their own systems based on its own platform and core technologies, and provide the deep learning platform to partners to build their own applications with the collected data. In terms of autonomous driving, SenseTime is based on original computer vision perception technology and self-developed FPGA computing platform that can quickly deploy deep learning model. Lane lines, driving areas, road information, pedestrians, vehicles and other information can be provided only through a single camera. At the same time, it can also provide customers with the multi-sensor fusion of autonomous driving solutions. In terms of the intelligent vehicle system, SenseTime has demonstrated its intelligent cockpit system based on SenseDrive1.0 architecture at the Automobile Shanghai in April. It hopes to create the first interactive method based on computer vision technology based on its own advantages. At this summit, SenseTime unveiled its overall SenseDrive2.0 architecture. It is worth mentioning that the SenseRover Pro/Mini, a self-driving car specially designed for children was also released. Autonomous driving underlying technology At present, there are two research and development directions for autonomous driving. One is autonomous driving using lidar based sensors, which rely heavily on three-dimensional, high-precision maps. The other is based on the camera, and auxiliary radar sensor, so that the overall cost can be done relatively low, in line with the needs of the home market. In a sense, SenseTime belongs to the latter group. Its committed to promoting mass production of autonomous driving solutions based on the technical reserve in the direction of computer vision AI technology. In terms of deep learning, SenseTime constructs the Parrots Primitive Library for deep learning and deploys SenseParrots, a platform for deep learning. The training model is the core element of deep learning, and its training platform is mainly based on GPU, and its goal is distributed, high-performance and extensible. Therefore, SenseTime has set up a supercomputing centre in Beijing, Shanghai and other places, with more than 15,000 GPUs to support algorithm training, including 17 clusters of GPUs, each cluster doing specific training for a vertical algorithm field. The largest cluster is close to 2000 CPUs. The powerful computing power brought by supercenters is one of the core elements that support the continuous iteration of artificial intelligence algorithms. With the popularization and development of autonomous driving, more and more companies begin to attach importance to the role of AI in the underlying architecture of autonomous driving, especially in the aspect of deep learning. Israeli vision systems company Mobileye, which previously used traditional algorithms for autonomous driving, has started to use deep learning algorithms in recent years, and their performance has improved significantly. XU Li (徐立), founder and CEO of SenseTime, said that "We only do algorithmic optimization and structural innovation. The data side is working with OEMs and the hardware side is working with Nvidia.“ In terms of vision, XU said SenseTime has summarized more than 20 vision algorithm modules, including front-end camera, camera imaging, calibration, backside recognition, understanding, scenarized and final path planning. These modules can be embedded in existing ADAS and provide the core technical support for future autonomous driving. Mass-produced self-driving solutions SHI Jianping (石建萍), research and development director of SenseTime, said that in the past three years, the company has mainly made efforts in perception technology, self-developed FPGA platform and so on. “SenseTime's new autonomous/assisted driving solution is based on original computer vision perception technology that provides lane lines, driving areas, road information, pedestrians, vehicles and other information through a single camera. It also promises to drive mass production by reducing the cost of building self-driving solutions.” SHI said. In addition, the SenseTime autonomous driving solution is equipped with an FPGA computing platform developed by the user to quickly deploy the deep learning model, which has strong adaptability and can be quickly extended to vehicles of different systems. Since entering the field of autonomous driving in 2017, SenseTime first worked with OEM Honda to jointly develop L4 autonomous driving solutions suitable for passenger car scenarios. The scheme can also be combined with a variety of different sensors to support technical functions such as perception, analysis and prediction, decision planning and control, urban-level 3d map reconstruction and high-precision positioning capability of unmanned vehicles, so as to realize a multi-level autonomous driving system. In 2018, SenseTime completed the implementation of no-takeover autonomous driving in semi-open venues in Hangzhou and Shanghai. In 2019, "AI self-driving park" will be launched in Japan, which will be used for the development and testing of self-driving cars and open to the public. In addition, SHI said the company will have more landing technologies for autonomous driving this year. If their deep learning network algorithm has a disruptive innovation, even though it's been in the car business for a short time, transcendence in the field of autonomous driving is not impossible for SenseTime.

Analysis EO
Analysis · 2
report
Analysis EO
May 17, 2019 05:11 pm ·

SenseTime Needs Overhaul of Sprawling Business Interests

With less than five years of operating history, SenseTime has closed more than 10 rounds of financing by well-known investors from around the world, Tiger Global, Silver Lake, Fidelity, SoftBank, Qualcomm, Alibaba and Temasek Holdings. The valuation has skyrocketed after rounds of funding, leading to both positive perceptions and questions on the feasibility of SenseTime’s business model: The team consists of scientists and professors intent on publishing academic papers and participating in global AI competition, but it is weak on monetization, and its vaunted “AI for everything” strategy seems impractical at this stage. Comparatively, other three Computer Vision startups in China, Megvii (originally known as Face++), CloudWalk, and Yitu Technology remain much more “low-profile” in public, though they have also drawn millions from investors. For example, Megvii, another Alibaba-backed startup and the biggest competitor to SenseTime, completed its Series D funding round last year, but chose to postpone the announcement until May 2019. With the intention of "cooling the market”, the company repeatedly emphasized that news of a new funding round did not carry the same weight as a major technological breakthrough. An AI business, just like companies in other sectors, should have its own business cycles. Moreover, the rest of the three startups are narrowing their focus to merely a few industries, convinced that startups worth less than USD 10 billion should focus on specific verticals instead of doing everything. In addition to using facial recognition technology in public security and consumer-oriented mobile apps, Megvii mainly focuses on logistics; Yitu specializes in healthcare; CloudWalk has chosen the financial sector. Different from its peers, SenseTime has been exploring over 10 verticals and is “impatiently” looking to apply AI to more industries, including self-driving, surgery and diagnostics, augmented reality, chip technology, AI education and so on. During the 2019 Artificial Intelligence Summit on May 15, SenseTime released 11 products across five industry verticals, the largest number of new products unveiled at a single event since the company was established.  How SenseTime turned into the most expensive AI startup Despite the mixed comments from different groups, we believe SenseTime’s rapid ascent into the world’s highest-valued AI startup is driven by the factors listed below, 1) The “best-in-kind” team. The company has grown to 3,000 employees now, doubling the number from a year ago. The team consists of PhD holders and college professors from MIT, Chinese University of Hong Kong, Tsinghua University and Peking University as well as tech experts from Microsoft, Google, Lenovo and Baidu. Everyone in the core development team has over 20 years of research experience. According to the founder TANG Xiaoou (汤晓欧), “There are less than 200 people with PhD on deep learning technology in China and SenseTime employs 120 of them. If SenseTime cannot offer the technology, the rest of the players have even smaller chance to do so.” 2) Greater government support. SenseTime has become the fifth National Open Innovation Platform for Next Generation Artificial Intelligence, after Alibaba, Baidu, Tencent and IFLYTEK. The Chinese government reportedly backs Baidu on autonomous driving, Alibaba Cloud on city brain, Tencent on healthcare medical images, IFLYTEK on Natural Language Processing (NLP) and SenseTime on Computer Vision technology. While the other four companies are tech giants with stable revenue streams, SenseTime is a new entrant to the list. We expect the policy to give SenseTime greater access to subsidies and increase the likelihood for it to land government contracts and hire more AI talents than its competitors. 3) The “AI for everything” plan underpinned by financial clout. Relative to its peers, SenseTime has the most funds available to deploy significant resources and support its aggressive expansion strategy. According to the company's vision, if it does not take initiatives now, the hotly contested market will be quickly taken over by emerging players in the next few years. Even if SenseTime might not have a substantial impact on certain areas, the company intends to build industry connections and brand awareness. When technology and solutions become more mature, corporate clients are more likely to choose SenseTime over vendors with less visibility. 4) Being a frontrunner on revenue and profitability. Though the specific revenue or net income figures are not disclosed, our industry researchers confirmed that in 2017, SenseTime posted revenue of around CNY 700 million and became profitable, just like Megvii. In 2018, sales reached CNY 7 billion, well exceeding all competitors'. Most of the revenue has come from government clients, with whom SenseTime could leverage its strong connections. Among all sectors that SenseTime is currently exploring, we highlight education, Smart City and entertainment, areas we think SenseTime has competitive advantages in. Governments are chasing AI education solutions Recently, SenseTime introduced its internally designed AI textbooks, “Fundamentals of Artificial Intelligence”, and distributed them to around 40 high schools in Beijing, Shanghai, Shenzhen and Shanxi. To spark students’ interests, its platform “SenseStudy” offers over 200 hours of AI and programming courses. Students can experience and learn about facial recognition, gesture recognition, augmented reality and many other trending technologies.  It is the governments who are spending millions of dollars on SenseTime’s education solutions. The reason why the government introduced AI to K12 education was that while US has been promoting STEM education (Science, Technology, Engineering and Math) for decades, China suffers a shortage of engineers, physicians and data scientists. The edge in scientific and technological innovation has become only more important amid the current escalation of US-China trade war. We believe that with SenseTime’s strategic partnership with the government-affiliated China Education Association, it is relatively easy for the business to secure additional orders and supply more schools with AI textbooks, training and lab equipment. In this sector, the path to monetization is clear and will likely be sustainable in the long run. Smart City is a big pie with a small threat from new entrants Visual surveillance systems installed in venues ranging from train stations to airports regularly receive technological updates. Besides public security, the scope of Smart City is even broader, including vehicle license plate recognition, traffic data identification, crowd flow analysis and so on. Cities and provinces are delegating Smart City projects to different vendors, who win contracts through competitive bidding. While 2017/2018 was the time for AI vendors to tender bids for contracts, companies which have won deals are in the process of fulfilling their contractual obligations and providing services now. Since public security, which makes up a big portion of Smart City projects, involves considerable sensitive information, governments are unlikely to grant such projects to a provider without a trusted relationship. Therefore, in a market worth around USD 80 billion with double-digit growth rates, there are limited opportunities for new entrants. SenseTime has developed particularly strong relationships with both domestic and foreign governments, demonstrated by 1) it won contracts with over 100 cities, nearly a third of all cities in China, and has installed over 100,000 smart cameras. 2) it has built research centers globally and has undertaken strategic cooperation with Southeast Asian countries, selling Smart City solutions to Thailand, Malaysia and Cambodia. It operates a heavier business model than its peers, who develop their businesses remotely or have yet to expand overseas.  AR brings higher user engagement to the entertainment sector SenseTime helps mobile game apps, social video apps, and other user-generated content (UGC) and other video platforms blending innovative human-computer interaction with user-generated content via augmented reality (AR) technology. For example, Tencent's first AR mobile game "catching the demon” was launched last month; it adopted a cutting-edge mix of computer vision, depth detection and real-time object recognition, provided by SenseTime. On account of declining user engagement and rising user churn rates, mobile app providers are looking for ways to increase user “stickiness”. If an app operator starts to use AR technology to attract users, its competitors are likely to make the same move or it risks driving its users to other platforms and losing market share. Downside risks Under this business model, downside risks are easily identified. Trying to cover many sectors is likely to distract SenseTime from concentrating resources to develop industry expertise - the real moat - and may lead to a diminished role for the company in certain industries, such as autonomous driving, which requires years of development. In addition to the government clients, SenseTime claims to have teamed up with over 700 corporations in China and overseas, such as Qualcomm, Nvidia, Huawei, Xiaomi and JD.com. However, the value that SenseTime creates for its enterprise clients may not directly be reflected in their financial statements. For example, Meitu, a leading photo-editing app which heavily uses SenseTime’s facial recognition technology to let users beautify their photos, failed to prevent its shares plummeting despite using AI and even blockchain technologies. It is challenging for SenseTime to drive profitability by serving only one or two industry leaders who have deeper pockets.

Analysis EO
Analysis · 2
report
Analysis EO
May 15, 2019 07:45 pm ·

China's AI Giant SenseTime Launches 11 New AI Products At Summit

SenseTime held the Second Artificial Intelligence Summit on May 15 and launched and displayed series of AI products and solutions for various fields, including smart city, smart healthcare, smart retail, AI education and augmented reality (AR). In total, SenseTime launched 11 new AI products at this summit. Let's take a look at what they are. Smart city integrated solutions SenseTime launched four products in the category of smart city integrated solutions: SenseFoundry 2.0, SenseNebula, Sense Pass and SenseStudio. SenseFoundry 2.0 is a computer vision analysis cloud platform for city security management. The first generation, SenseFoundry 1.0, was debuted last year, which is capable of processing more than 100,000 way of signals at time. And it can analyze over 100 billion unstructured and structured features. Combined with deep learning, SenseFoundry 1.0 has multiple functions like, among others, face recognition, vehicle recognition, passenger recognition, crowd analysis and events detection. Compared with the first generation, SenseFoundry 2.0 boasts upgrades including high precision algorithm set, algorithm models for various application scenarios and multi-dimensional analysis engine. Among them, high precision algorithm set consists of several new algorithms, such as advanced face and human body recognition algorithm, enhanced structuring algorithm and remote sensing analysis algorithm. SenseNebula is a smart device, which is designed to realize edge computing function. It can be connected with many kinds of front-end devices when needed, like camera, capture machine, access control device, visitor register machine and so on. Every interface on SenseNebula has access to different algorithm models for different application scenarios. SensePass is a front-end product used to enable human body recognition with an error rate of only one in 10,000. It supports 20,000 human body offline databases, which means it can be equipped without installing complicated background system. Besides, it’s only 15mm thick and stands in harmony with the surrounding environment. Sense Studio is an open platform that contains multiple AI applications, supporting transportation, finance, school, hotel and scenic spot. Users can choose the applications they need and use them to upgrade their business without much obstacles. SenseTime has provided its smart city solutions to almost 100 cities from 31 provinces in China, in which SenseTime has deployed a fleet of over 100,000 smart cameras. In one city of Eastern China, SenseTime builds the biggest smart city system globally with 20,000 smart cameras and 2,000 GPU servers. AI education solutions In the field of education, SenseTime introduced its AI textbooks for high school students. This is a product of joint efforts by teachers, experts from SenseTime, the Commercial Press, East China Normal University Press and three other educational institutions. In addition, SenseTime plans to design a series of AI textbooks for middle school this year. Based on deep learning platform “SenseParrots”, SenseTime launched AI education experimental platform “SenseStudy”. SenseStudy offers over 200 hours of AI and coding courses. Students can do some practice and experience on face recognition, gesture recognition, augmented reality and many other technologies. SenseTime also launched two robot products for AI education: SenseRover Mini and SenseRover Pro. SenseRover Mini is a car programmed to coordinate with AI courses and it is mainly used by freshman and sophomore students. Sense Rover Pro is a self-driving car, with automatic cruising, lane detection and many other functions. Users can type codes on the SenseStudy to realize motor control and conduct some experiments on the sensors. In addition, SenseTime organizes teacher trainings, AI speeches and summer camps, as well as some seminars, in order to cultivate more qualified teachers and talents proficient in AI. At the summit, SenseTime announced a plan to set up an AI education laboratory with Nanjing Normal University, and they will jointly develop advanced AI education methods. Smart healthcare solutions Instead of developing AI-assisted imaging diagnosis system, SenseTime debuted SenseCare, a smart diagnosis and treatment platform, which can serve the needs of hospital and process most types of data. SenseCare has capability of high concurrency of 3D rendering and clinical application. It can show data to patients or users through a  visualized interface. At the same time, all the data are stored in hospital’s servers to ensure the data security. This product has been adopted in a hospital in Shanghai, specifically, in the oncology department, to help doctors diagnose cancers. Smart retail solutions Also unveiled at the summit is SenseGO, a smart retail platform. Combined with smart camera, SenseGO can collect and analyze customer behaviors offline. SenseTime also developed shopping windows with some intelligent functions to attract customers’ attention. SenseGO has been adopted in a supermarket of 200,000 square meters, and it can serve up to 50,000 customers at one time. AR solutions SenseTime upgraded its AR solutions with new function and business cooperation. SenseAR, the function of effects for videos, live broadcast and images. SenseAR can recognize and identify people’s faces, hands and other parts of body. Based on image recognition, SenseAR can provide effect solutions. Besides, SenseTime announced that it will partner with AR glass manufacturers to jointly develop AR glasses . SenseTime also revealed that it wound cooperate with Qualcomm in making chips and 5G SDM8150. XU Li (徐立), CEO of SenseTime, said at the summit that The mission of SenseTime is to apply AI to every part of people's life and cut production costs. SenseTime will develop more AI products and supply them  to more industries.

Analysis EO
Analysis · 2
Analysis EO
Jan 12, 2019 05:34 pm ·

10 Innovative Chinese Startups to Watch at CES 2019

There once was a joke that the annual gadget show in Las Vegas CES in fact stands for the "Chinese Electronics Show," because there were 1,551 Chinese companies registered to the CES's exhibit in 2018, accounting for 33% of all exhibitors. But the number of Chinese exhibitors is 1,211 at CES 2019, with a 20% decline from last year, according to the Consumer Technology Association, which organized the show. And that is as it should be as the US-China trade war intensifies and the world's two biggest economies are fighting for control over a range of frontier technologies such as artificial intelligence and 5G mobile networks, which are also part of the focus of this year's CES. China's telecommunications equipment giant Huawei hasn't sent any executives to the consumer electronics show, comparing to last year when Huawei CEO Richard Yu announced the Mate 10 Pro to the US market. Another telecommunications equipment company ZTE Corp., whose name is in the exhibitor dictionary, didn't appear in the CES this year, giving no explanation as to why it is missing the show for the first time since 2011. With the absence of giants like Huawei and ZTE, there still are a great number of Chinese startups in the show, they have brought the world with their newest achievements of frontier technologies such as artificial intelligence and autonomous driving. And here are 10 of them you should keep an eye on at the CES 2019. ForwardX Robotics (灵动科技) Self-driving technology and algorithms are not for cars only, they can also appear in items in your life that you will never expect, for example, your suitcase. At this year's CES, Chinese autonomous-focused company ForwardX Robotics demonstrated its latest product Ovis Suitcase, an autonomous piece of suitcase that is powered by artificial intelligence and self-driving technologies. Unlike standard suitcase, the Ovis Suitcase uses self-driving algorithms to side-follow users anywhere, including hotels, airports and even city streets. This eye-catching suitcase uses deep learning-based computer vision for high-level understanding, VSLAM (visual simultaneous localization and mapping) positioning, reinforcement learning-based navigation and avoidance, and automatic control technology. These technologies allow its products to have a "smart brain" with perception, cognition, judgment and decision-making capabilities, according to the company's website. Founded in 2016, ForwardX Robotics is an AI company based in both China and the US. The company committed to building the "robotic brain" of the AI era, using deep learning to redesign robots. The company has introduced both an enterprise product visual-based Autonomous-Mobile-Robot (AMR) and the consumer product Ovis Suitcase. Byton (拜腾) Founded in 2016, Chinese electric vehicle manufacturer Byton has already a Unicorn. The company raised altogether $500 million in Series B funding at a valuation of $1.9 billion. Byton is also seeking to raise at least $500 million to finance growth, valuing the nearly three-year-old company at more than $4 billion, Reuters reported on Tuesday. At CES this year, Byton demonstrated its concept car M-Byte, which features with a 48-inch-wide central console video display. Byton also collaborates with Amazon to provide the car voice control by Alexa. The car Byton showed at CES is still concept version, but the company has a production facility in Nanjing, China, and says that it has built about 100 prototype vehicles so far. And According to Reuters's report, Byton is looking for foreign investors in its latest fundraising and the fund will mainly be used to finance the mass production of M-Byte. See also: BYTON Finalized The Design to be Mass Produced, 48-inch Screen Retained YI Tunnel (图灵通诺) If you are familiar with Amazon's cashierless store Amazon Go, you'll easily understand what this Chinese startup YI Tunnel is trying to do. The company has already opened the first cashierless store in China based on the same vision recognition technology with Amazon Go, the only difference is that it costs only 5% of Amazon Go. The company's vision is to bring technology and commerce together by leveraging the power of AI. The company uses AI to digitize the elements involved in the retail industry. YI Tunnel has already deployed its cashierless store in different countries, such as Etisalat in Dubai, PepsiCo in New York City, and SK in Seoul. In addition to cashierless stores, YI Tunnel also has developed a whole set of AI-driven products for the retail industry, including an AI autonomous checkout machine, a big data analytics system for SKU (Stock Keeping Unit) and tools for supply chain forecasting. RoboSense (速腾聚创) Chinese LiDAR perception technology solutions provider Robosense demonstrated at CES 2019 an upgraded version of its MEMS (Microelectromechanical Systems) solid-state LiDAR, the RS-LiDAR-M1, which is designed for the mass production of autonomous vehicles. At the CES last year, Robosense had already announced the first generation of its MEMS solid-state LiDAR. The product was initially applied to Alibaba's autonomous logistics vehicle. The upgraded LiDAR provides an increased horizontal field of view reaching a 120° field of view, so that no more than 5 RS-LiDAR-M1s will be enough to cover the 360° field of view, and realize Level 5 (SAE) automatic driving-full driverless driving. Other companies to watch: Keep (Keep健身) Keep is a sport-tech brand born in China, the company's mobile fitness application that provides online fitness training courses attracted over 100 million users. Keep closed a $126 million series D funding led by Goldman Sachs on June 10, 2018. The company is dedicated to providing a one-stop service towards home training, running, cycling, social network, nutrition guide and sporting goods. The online social fitness app even opened its own offline fitness centers called the Keepland. Horizon Robotics (地平线机器人) China-based innovative company Horizon Robotics introduced its award-winning product Matrix Autonomous Driving Computing Platform (MADCP) at the CES 2019. The product was named the CES 2019 Innovation Awards Honoree in the Vehicle Intelligence and Self-driving Technology Category. Matrix is a self-developed autonomous driving computing platform, based on the BPU (Brain Processing Unit) 2.0, a proprietary and high-performance AI processor architecture independently developed by Horizon to empower L3 and L4 autonomous driving. The product is now proving for large autonomous driving suppliers in mass production, the company claimed. Horizon Robotics is one of China’s leading designers of artificial intelligence chips. The company raised up to $1 billion in a funding round and valued it at between $3 billion and $4 billion. SoundAI (声智科技) SoundAI has presented an intelligent interaction system SoundAI Azero and smart home one-stop solutions – MiniSpot and MiniBox at the CES 2019. SoundAI Azero is a whole-chain far-field intelligent interaction system developed by SoundAI, with far-field speech interaction and far-field real-time communication as technology core. It is compatible with popular intelligent chip architecture, board assembly and acoustic design, with built-in lifestyle service, information query, IoT control, entertaining content, utility tools, personal assistant, chatbot, etc. It also supports third-party skillsets development as need arises. See also: 2019 CES SoundAI’s New Voice Interaction Technology Overseas Debut SenseTime (商汤科技) SenseTime is a China-based AI unicorn focused on computer vision and deep learning. The company has raised a total financing of more than US$1.6 billion and is valued over US$4.5 billion. SenseTime has developed a deep learning platform, supercomputing centers, and a range of AI technologies such as face recognition, image recognition, object recognition, text recognition, medical image analysis, video analysis, autonomous driving, and remote sensing. In doing so, SenseTime has become one of China’s largest AI algorithm providers. See Also: To Build Smart Retail, SenseTime and Suning Forms Strategic Alliance at CES 2019 TuSimple (图森未来) TuSimple (图森未来), a China-based autonomous driving company with an R&D facility in San Diego, announced that the company will have 40 trucks in fully-autonomous operation by June in U.S. The company is displaying a Navistar International LT semi-truck at CES 2019. TuSimple is an innovative technology company that is devoted to the development and application of artificial intelligence and computer vision technologies. See also: TuSimple Displays Navistar Semi-truck at CES 2019 and Announces Business Expansion Plan Tuya Smart (涂鸦智能) Tuya Smart, a China-based "AI+IoT" developer platform that makes plugins to enable manufacturers to transform ordinary products into smart devices, has announced two new solutions at CES 2019. The AI Video Doorbell and the AI Pad, are next-level security solutions for the home and office that incorporate artificial intelligence and the Internet of Things. See also: China-based Tuya Smart AloT Platform is set to debut products at CES

Analysis EO
Analysis · 2
Analysis EO
Dec 1, 2018 05:14 pm ·

SenseTime Embraces Cooperation as an AI Plarform

Beijing, Nov 29, 2018 /EqualOcean/ - EO Innovators Annual Conference & the 4th Innovation Awards Ceremony was held by EqualOcean (亿欧国际) and sister website iyiou.com (亿欧公司) in China World Summit Wing. In a panel discussion themed with “Cutting-edge Technologies & Intelligent Industries”, CHEN Yuheng (陈宇恒), co-founder and senior engineer of SenseTime (商汤科技) showed up and shared his thoughts on company positioning. Panel Discussion of Cutting-edge Technologies & Intelligent Industries “For SenseTime positioned as a platform, it is necessary to break the boundary between research labs and industries, to cooperate with more partners and establish systems that can be used in real industrial scenarios.” CHEN Yuheng said. Furthermore, CHEN Yuheng shared the main business practices of SenseTime and thoughts about the future. The following is the collection of highlights of CHEN’s speeches during the panel discussion: GAO Xinin (高欣欣, moderator of the discussion, CEO of Jiangmen Capital): What industries do you believe will be better enhanced by technologies in the future? CHEN Yuheng: SenseTime is now focusing on several products, the most important one is Computer Vision based Intelligent Security, as well as the Smart City; the second is about Smartphone and Intelligent Entertainment, including face identification, beauty camera, etc., which applications were provided to smartphone manufacturers. As to the future, I mean in the next few years, SenseTime chose to believe in opportunities of AI-enhanced Education, which is also quite meaningful. So far, we had published several AI textbooks for middle school students, and also offered the experimental platform to try out the students’ new ideas. GAO Xinxin: Where’s the boundary between technology platform and its partners? What should the platform do, and what should be for partners? How to accelerate industrial upgrade with the power of ecology? CHEN Yuheng: Most of the time, AI algorithms trained in a lab with simulated data could not be easily applicable in the real production environment. Thus, as a platform, it is necessary to break the boundary between research labs and industries, to cooperate with more partners and establish systems that can be used in real industrial scenarios. Let me give you two examples: Algorithms for applications like Face Identification and Beauty Camera should be deployed in the computing chips on smartphones manufactured by Oppo or Vivo, etc., and the persistent optimization of user experience needs the actual behavioral data from the users. Participants in Intelligent Security includes system integrator companies, camera manufacturers, and algorithms companies like SenseTime. Different companies should work together to optimize the product, to better meet the needs of end-clients. For instance, the product should able to help the police to investigate, or to help people find their lost family members, etc. The cooperation between the technological platforms and industries will undoubtedly generate actual values, however, the boundaries of trade secrets or private data should still exist. Besides CHEN Yuheng of SenseTime, the other two participants of the panel discussion are WANG Tianmiao (王田苗), the Ph.D. supervisor in Beihang University (北京航天航空大学), and WANG Zhuoran (王卓然), the founder and CEO of trio.ai (三角兽). - Author: ZHANG Fan; Write to ZHANG Fan at ZhangFan@EqualOcean.com

Analysis EO
Analysis · 2
Analysis EO
Nov 20, 2018 11:47 pm ·

Sensetime is about to face the revenue growth challenge

Sensetime (商汤科技) is definitely a star company among the AI technology startups, and even unicorns in China. Founded on Nov 14, 2014, Sensetime had already raised a total funding amount of approximately USD 2.67 billion (estimated according to public data). Wrapped up with the latest funding round with the amount of USD 1 billion in Sep 2018, from SB China Capital, this AI tech company was valued at USD 6 billion, to be the most valuable Computer Vision unicorn in China, and this valuation actually exceeded the market capitalization threshold of the top 500 listed companies in China, 30 billion yuan, according to the data of 2017. China's 5th national AI open innovation platform Sep 20, 2018, the Ministry of Science and Technology of PRC officially announced relying on Sensetime to build a new generation of AI innovation open platform for Computer Vision. Since then, Sensetime has become the 5th national artificial intelligence innovation open platform, after Baidu(the autonomous driving open platform), Alibaba(the urban brain open platform), Tencent(medical imaging open platform), and iFlytek(intelligent voice open platform). The announcement could be recognized as the symbol of government's recognition, which will project positive influences on the cooperation with the government. Plus, the signal gives the market some confidences, although it will not be the key consideration factor for Sensetime's clients. By focusing on innovative computer vision and deep learning technologies, Sensetime tried to empower its skills in a variety of industries, including Intelligent Security, Terminal, Entertainment, Finance, Business, Remote Sensing, Robot Sensing and Control, Autonomous Driving, etc. The cross-industry technical applicability is the basis for Sensetime's responsibility as the builder of the “national open platform”. Sensetime had its technology advantages, while which was not invincible Sensetime was founded by TANG Xiaoou (汤晓鸥), who also founded the Multimedia Laboratory at the Chinese University of Hong Kong, which was the only one of the world's top ten artificial intelligence labs located in Asia and the R&D team had published more than 400 computer-vision-related papers on international leading academic journals and conferences. In 2014, Sensetime released its face recognition algorithm DeepID, which was claimed to be the world's first algorithm that could exceed human being's accuracy of face recognition. However, on Nov 15 2018, the most updated Face Recognition Vendor Test Report released by NIST (National Institute of Standards and Technology) showed that Yitu Technology's two algorithms achieved the best performance and defeated Sensetime - the new entrant of this test. Additional algorithms not listed in the leaderboard can be found in Table 2 of the latest FRVT report.  Although this test result could only be a reference material for the industry, it still indicated that Sensetime's technologies and algorithms are not invincible as it claimed. Focusing on revenue would be more important for China's AI startups in 2018 Recently, while publishing his new book "AI·Future", LI Kaifu (李开复) had said that he thought that the concept of AI was way too popular in China, and many companies used "AI" just for raising money, which generated uncountable bubbles in the industry. "In the past three or four months, AI companies' valuation has dropped by 20%-30%. If it continues to drop another 20%-30%, the valuation would be reasonable, and we could start to increase investment", LI said. The industry now commonly believed that the technology itself is not the key indicator anymore (which was indeed the key in 2017), and what matters now would be the practical application of technologies in varied industries. Sensetime did an outstanding job in 2017 with an alleged revenue of USD 100 million and expected to triple its revenue this year, to $300 million, according to an article on Bloomberg. (The company declined to comment about this number) According to EqualOcean's speculation, Sensetime, cooperated with the government, should have already achieved revenue goal from the security industry, which could contribute 70% of its total revenue and the rest 30% would be contributed by its cooperation with smartphone and software companies. However, the biggest challenge for Sensetime would not be generating income, but to increase revenue rapidly. In the Security industry, traditional players like Hikvision and Dahua Technology would not be easy to compete with, and as for the internet and computing device industry, clients themselves have good technical accumulations and always try to get rid of their technology suppliers. On the other hand, the competition among the startups cannot be ignored. Nowadays, there are four AI unicorns related to Computer Vision in China, which is also believed to be unreasonable in LI Kaifu's opinion before in 2017. Although Sensetime currently has a relatively dominant position, it is still far away from where the company could be safe to loose its nerve. - Author: ZHANG Fan; Write to ZHANG Fan at ZhangFan@EqualOcean.com