Computer Vision Community Develops JOYCE, a Humanoid Robot with Intelligent Vision using Data-in-Picture Technology

Published  September 25, 2020   0
Humanoid Robot JOYCE by Computer Vision Community

JOYCE from Immervision is the first humanoid robot developed by the computer vision community to help machines gain human-like perception and beyond. This humanoid robot JOYCE invites the community to help her gain a better understanding of her environment by upgrading her optics, sensors, and AI algorithms to further advance computer vision technologies.

The company has also come up with the JOYCE development kit for engineers and AI developers. The JOYCE inbox development kit is equipped with three ultra-wide-angle perimorph cameras calibrated in a way to provide 2D hemisphere, 3D stereoscopic hemisphere, or even full 360 x 360 spherical capture and also for viewing the environment. 

JOYCE will be able to enrich her video frames with the help of data-in-picture technology. The data from a wide array of sensors provides contextual information to the neural network, AI, also computer vision, and SLAM algorithm that helps in increasing the robot’s visual perception. It will be streaming live so that people can not only follow her evolution capabilities but can also look through her eyes when she travels all around the world, perform skydives, and also visits computer vision lab or any business conference.

The main aim to build the humanoid robot, JOYCE is to accelerate the emergence of the new innovative solutions and build the next generation of intelligent systems especially for a wide range of the industry devices. JOYCE aims to address various challenging situations. Enhancing the performance of smart home devices, providing improvement in the optics technologies for driver safety in assisted driving, or in autonomous cars are its potential use cases. Other than that, it will also help improve firefighting ability for detecting people and objects above the tree and also through the smoke. Adding to the list are providing medical diagnostic for identifying cancer tumors or various conditions in CT scans and helping in identifying early signs of crop diseases, and many more.

Key Features of Joyce Camera

  • 3 x 187° panomorph lenses: These provide side-by-side stereoscopic wide-angle capture and full 360° degree surround capture
  • Auto Exposure
  • Auto White Balance
  • Image quality, resolution, and frame rate control
  • Output format (YUV, MoTon JPEG)
  • Connectivity: USB3 (UVC Compliant)
  • Includes Inertial Motion Unit (Gyroscope) and GPS Sensor