Work in Progress: Biologically Inspired Dynamic Task Prioritization in Computer Vision Systems

This paper describes our work in developing a computer vision system that more closely mimics the way the human vision system operates. Autonomous vision systems are crucial for modern vehicle safety, demanding real-time object detection and prioritization without external computational resources. Current systems struggle with this on-board prioritization. This paper presents a novel approach inspired by the human eye and brain, segmenting the video input into foveal and peripheral areas for specialized processing.