LOVOT, a family-type robot developed by robot venture GROOVE X Co., Ltd., has announced that it will be equipped with its own "FORXAI Imaging AI" skeleton detection technology developed by Konica Minolta Co., Ltd. that detects human posture with AI as a new feature of "LOVOT."
"FORXAI Imaging AI" skeleton detection technology is a technology that estimates what kind of posture is being taken by AI automatically detecting a person in an image and detecting that skeleton at high speed.
With this technology, when the LOVOT is looking at you in the distance, it improves the accuracy that the LOVOT will approach you when you crouch to match the level of your eye line, and the skinship with the LOVOT will be richer.
The update of this feature is scheduled for the end of 2022/5. Utilizing "FPGA" and embedded AIKonica Minolta has core technologies in the field of cameras, images derived from photographic technology, and optics, and is developing optimized and pre・learned deep learning models for object recognition and attitude recognition.In the field of manufacturing, we are developing embedded AI using FPGAs, which are integrated circuits that can change circuit configurations, and implementing deep learning technology into IoT devices (edge devices) such as network cameras.
GROOVE X has agreed to challenge the development of image IoT/AI technology to expand the expressive power of Konica Minolta and LOVOT, and has been promoting it since spring 2020. And this is the first new feature to further develop attachment formation with "LOVOT".
"FORXAI Imaging AI" skeleton detection technology, those who participate in the "Advance delivery tester" will be delivered sequentially after 2022/5/10 (Tue).The update is scheduled for the end of 2022/5.
Outline of "FORXAI Imaging AI" Skeleton Detection TechnologyKonica Minolta's "FORXAI Imaging AI" skeleton detection technology is a technology that estimates the posture of a person by automatically detecting a person in an image and detecting the skeleton at high speed.
Utilizing the image processing technology cultivated in information equipment and diagnostic imaging equipment, it not only achieves both high recognition accuracy and high-speed processing, but also realizes low power consumption and low cost through edge computing.
