On August 11, Xiaomi held a new fall conference, and Xiaomi’s first full-size humanoid bionic robot CyberOne officially unveiled. According to Officera, CyberOne is equipped with a Mi-Sense depth vision module designed by Xiaomi and developed in collaboration with Officera; combined with AI interaction algorithm, it not only has complete 3D space perception capability but also can realize character identity recognition, gesture recognition and expression recognition.
Xiaomi’s full-size humanoid bionic robot CyberOne is 177cm tall, weighs 52kg and is known by the stage name “Tie Da”. According to Xiaomi, “Tie Da” adopts a variety of technologies, and can sense 45 kinds of human semantic emotions and distinguish 85 kinds of environmental semantics; equipped with Xiaomi’s self-developed whole-body control algorithm, which can coordinate the movement of 21 joints; equipped with Mi Sense visual space system, which can reconstruct the real world in three dimensions; the whole body is driven by 5 kinds of joints, and the peak torque is 300Nm. Xiaomi said that the intelligent and mechanical capabilities of CyberOne are all self-researched by Xiaomi Robotics Lab. There is countless software, hardware and algorithm development work behind this.
Recently, Ovation has released its self-developed machine vision depth camera module, which mainly consists of iToF module, RGB module, and an optional IMU module. iToF technology is one of the mainstream 3D vision perception technologies nowadays. iToF technology is one of the mainstream 3D vision perception technologies, and Ovation develops 3D intelligent depth cameras based on iToF+RGB depth measurement technology.
In fact, as the eyes of the machine, ToF depth camera is one of the best options to enable intelligent devices to see the world and have the ability to sense and understand the environment. The depth camera can identify the 3D coordinate information of each point in the space within the field of view, which enables the computer to get the 3D data of space and recover the complete 3D world and realize various intelligent 3D positioning.
The 3D intelligent depth camera released by Officera consists of three main parts: iToF module, RGB module, and optional IMU module.
In terms of external interfaces, this intelligent depth camera has USB 3.0 Type-C type interface for depth, IR and RGB, IMU and other data transmission, USB 3.0 bandwidth can theoretically support the transmission of point cloud data, the function is more scalable. At the same time, considering the synchronization of other sensors and depth camera, the depth camera retains the synchronization signal interface as well as some GPIO interfaces and power supply interfaces, making it easier for users to integrate multiple sensors.
Ovation introduced that the intelligent depth camera has a powerful SoC platform, which integrates ISP, ARM CPU and embedded GPU and NPU IP, and can consider releasing the arithmetic power for users later, so that users can better utilize the arithmetic power of the depth camera. At present, the depth decoding occupies part of the GPU computing power and part of the CPU computing power, and the NPU computing power can be fully released later.
The released iToF-RGBD depth resolution supports up to VGA (640×480) with a frame rate of 30fps, and will open higher frame rate mode and support binning mode later.
The depth FoV is 76° horizontally and 61° vertically with a laser wavelength of 940nm, while the RGB uses a 1.3-megapixel sensor with a global shutter and an FoV of 93° horizontally and 82° vertically.
Regarding the depth measurement performance, it currently supports the measurement range from 0.1m to 5m indoors (test environment: indoor illumination not more than 1w Lux, 90% reflectivity white wall) and 0.1 to 3m outdoors (test environment: outdoor not more than 5W Lux, 90% reflectivity subject), with an accuracy of up to 1% in the measurement range.
In the field of face recognition, it is mainly used for face payment, authentication face access control, face attendance, etc.; in the field of robotics, it is mainly used for obstacle avoidance, SLAM, AI recognition, etc., in addition to human body modeling and other scenarios; in the field of 3D scanning, it is mainly used for indoor scanning modeling; in the industrial field, it is mainly used for industrial automation parts scanning detection sorting. r / ar, access control, intelligent transportation and many other One of bottlenecks in the development of the field.
