Himax Technologies will unveil the next-generation version of its WiseEye AI processor for battery-powered tinyML products and reference designs at next month’s Consumer Electronics Show.
TinyML, short for tiny machine learning, is an emerging field in technology focused on optimizing machine learning (ML) algorithms to work on low-power, low-resource hardware, such as microcontrollers and small processors. This approach enables the integration of intelligent features into tiny devices, including wearables, IoT devices, and various embedded systems, without relying on continuous cloud connectivity.
TinyML is characterized by its minimal energy consumption, making it suitable for battery-powered or energy-constrained environments. It enables a wide range of applications, from voice recognition and gesture control to environmental sensing and predictive maintenance, by bringing AI capabilities directly to the edge, closer to where data is generated. This not only reduces latency and power consumption but also addresses privacy concerns by processing data locally.
Himax says that its WE2 processor delivers 32 times faster inference speed for AI, and more efficient power consumption compared to its predecessor. At CES 2024, Himax will showcase a range of applications of the WE2, including palm vein authentication, facial recognition, and object detection, among others. Himax will also provide a live demonstration of facial mesh analysis, which normally requires dedicated graphics hardware.
Himax Technologies, Inc. is a leading global fabless semiconductor solution provider, primarily known for its advanced display imaging processing technologies. Founded in 2001 and headquartered in Tainan, Taiwan, Himax specializes in the development and manufacturing of display drivers, timing controllers, and other semiconductor products, but is also increasingly known for its contributions in tinyML visual-AI and optical technologies. Its first-generation WiseEye AI processor was used in a series of Dell premium notebooks, among other endpoint applications.
–
(Originally published on FindBiometrics)
Follow Us