This page was machine-translated and may differ from the original. View original

ARM Computer Vision Calculates Calories 'Right' Even with Just Volume Detected

Google 우선 소스Published2017.06.28 14:33

ARM Tech Forum Introduces Latest CPUs and GPUs to Lead the IoT and AI Era
DynamIQ Technology, Computer Vision, and Wireless Communication Technology Announcement

ARM, a semiconductor design intellectual property (IP) company, held its annual event, 'ARM Tech Forum 2017,' at the L Tower in Yangjae-dong, Seoul, on Wednesday, the 28th.

At this ARM Tech Forum event, various solutions were introduced, centering on the latest CPU and GPU products supporting the newly announced DynamIQ technology. During the presentation, ARM’s Chris Turner introduced premium mobile experiences through the ARM Cortex-A73, and Anand Patel presented on the topic of 'ARM GPU technology'.
ARM Tech Forum, ARM's annual event

In addition, ARM’s Steve Steele introduced next-generation image processing technology for automobiles while presenting ARM processors that drive automotive innovation.

In addition to this, there were presentations on topics such as computer vision, mbed updates, and recently emerging wireless communication technologies like LPWAN/NB-IoT and BLE.
Computer vision implemented based on ARM NEON and Mali GPU. Measures calories by measuring volume.

On this day, ARM demonstrated an application that implements computer vision and machine learning on a mobile device. The real-time calorie meter with applied artificial intelligence was implemented based on ARM NEON and Mali GPU.

"Computer vision capable of processing machine learning functions at the edge can calculate calories by measuring the volume of objects," an ARM official said. "This is processed at the CPU and GPU levels of the device rather than connecting to a server for analysis, enabling real-time processing and analysis."

ARM is currently partnering with various companies to develop applications utilizing computer vision solutions. It is expected to be used not only in video fields such as CCTV, but also in future sectors such as automotive and VR/AR.
본 기사에 대한 정정·반론·추후보도 청구는 보도 청구 안내를, 그간 게재된 보도문은 정정·반론보도 모아보기를 참고해 주세요.
신윤오 기자