Dynamic Vision Sensors (DVS) are asynchronous camera sensors that report changes in light intensity. These sensors have many advantages compared to traditional frame-based cameras, such as increased dynamic range, lower bandwidth requirements, and higher sampling frequency at a lower power cost.
In this work, we propose a low hardware complexity architecture for object detection with DVSs, by using an efficient preprocessing pipeline and a binary neural network for detection.
This project is part of my research at Samsung AI Center NY.