Purdue’s Computer Vision Device Mimics The Human Retina

Researchers at Purdue University have developed a prototype device that takes inspiration from light perception in retinal cells. The device triggers an electrochemical reaction in response to light, which creates a memory of the light information it has received. This memory could be used to reduce the amount of data that needs to be processed to understand a moving scene, making it a more energy-efficient and computationally efficient approach than traditional computer vision. The design may also prove useful in neuromorphic computing applications that follow the architecture of the human brain.