Showcase with Ecosystem Partners in tinyML Applications, Featuring High-Performance AI and Ultralow Power Consumption Across Diverse Domains
Himax Technologies, Inc., a leading supplier and fabless manufacturer of display drivers and other semiconductor products, announced that during CES 2024 the Company will unveil its new generation WiseEyeTM AI processor, WE2, with a diverse array of new sensor fusion use cases, encompassing image, video and audio. Himax has also teamed up with numerous industry-leading ecosystem partners in a joint showcase of its WiseEyeTM technology for a range of battery-powered tinyML products and reference designs, covering smart laptop, smart home, smart city, wearable device, and surveillance applications.
As demand for AI grows, the number of devices with powerful processors or cloud connections surges. Yet, the high-power requirements of these devices present challenges such as escalating costs, high power consumption and deployment limitations, along with the privacy concerns in cloud computing. Himax’s proprietary WiseEye AI solution features local inferencing tinyML vision AI with ultralow power, a compact form factor and built-in industrial grade security and cryptography engines. It is ideally suited for resource-constrained, battery-powered endpoint AI applications and seamlessly integrates into everyday life. Market adoption is swiftly expanding across diverse domains.
The Company’s first generation WiseEye AI processor, WE1, supports the mass production of a series of Dell’s premium notebook along with other endpoint AI applications. At CES 2024, Himax will reveal its next generation WE2 AI processor with 32 times faster inference speed AI, and 40% peak power saving compared to its predecessor, WE1. The WE2 exhibition will showcase applications such as palm vein authentication, face detection and recognition, facial landmarking, head pose estimation, hand detection, hand landmark detection, and object detection among others. A live demonstration of facial mesh analysis will also be presented, executing a sophisticated model, which traditionally requires discrete graphics hardware, to underscore its unique capabilities. This is a groundbreaking achievement, enabling professional-grade computer vision and perceptual AI functions to operate directly on miniature power-efficient endpoint devices with ultralow power consumption. Additionally, WE2 supports audio capabilities, including keyword spotting for voice control, seamlessly integrating with vision AI to deliver comprehensive sensor fusion capability.
Finally, the Company will also showcase a new line of production-ready WiseEye Modules, a versatile module which features an integrated computer vision hardware and software platform consisting of Himax’s low power CMOS image sensor, WE1 or WE2 AI processor, and Himax’s in-house DNN (Deep Neural Networks) models. The module is designed to reduce the entry barrier for development and enable a variety of applications, such as digit reading, people counting and object detection. It offers an effortless plug-and-play user experience due to its outstanding no-code/low-code AI, compact form factor and universal interface. At CES, numerous endpoint AI application collaborations with ecosystem partners will also be on display, such as Seeed Studio for a battery-powered endpoint AI vision processing module, DESMAN for a smart door lock, Murata for automatic meter reading (AMR) and Ganzin for eye tracking in mix-reality glasses, all illustrating the consolidated efforts between parties to broaden market reach.
SOURCE: Himax Technologies
