Feb 20, 2019, Xiaomi (小米) launched its new XiaoAI Touchscreen Speaker in a Beijing press, showing that the company has continued to stick to its AIoT (Artificial Intelligence of Things ) plan. Apart from the ambition to integrate all its smart home gadgets to be controlled from a central hub, the company also aims to deploy artificial intelligence technology to expand the methods of controlling the gadgets.
As a Hong Kong-listed company, Xiaomi is best known for its budget phones. Redmi, the new independent smartphone brand focused on the low-end market, has so far delivered over one million devices since its brand splitting on January 15, 2019. Xiaomi, as the company’s main brand, will provide Mid- to high-end devices that compete with Samsung and Apple (see more in this article).
To expand its artificial intelligence (AI) and IoT strategy, Xiaomi cooperated with Sweden-based retailing giant IKEA to offer smart lighting products last November, reported by Mobile World Live. Xiaomi sets to invest USD 1.5 billion in artificial intelligence and smart devices over the next five years, as the company announced one month ago.
Horizon Robotics (地平线机器人), a leading Chinese AI chipmaker, leveraged its Enhanced Speech Extraction in the design of the brand new speaker. Horizon has provided a complete front-end voice signal processing and wake-up phrase solution set for the new product.
Test results show ESE solution can extract voice signal efficiently in noise coming from television, refrigerator, washing machine, etc., meanwhile, with a high wake-up rate. To perfect user experience, the company also deployed depth algorithm optimization featuring front-end voice wake-up to guarantee a low error rate. The company also provides semantic comprehension and speech synthesis solutions apart from those mentioned.
The company’s founder and CEO YU Kai (余凯) claimed that the company’s mission is to leverage its technology for smart devices, or so-called Horizon Inside. The firm wants to make more edged artificial intelligence products with its advantage on voice and vision. Horizon has always enjoyed the label of Embedded AI Processor and Processor+ Algorithm+ Cloud Platform.
Apart from the smart home, Horizon voice solution set can be applied in vehicles. The company always consider combining algorithm and hardware together in every single product, from the chips Journey 1.0, Journey 2.0, to the platform Matrix (see more in this article). Voice algorithms developed by Horizon are perfect to achieve low power, low cost and can be easily integrated. The algorithms were also tested in Horizon’s first generation of chip successfully.