Event Intensity Decay with Event Cameras for Efficient Object Detection

Abstract

In this paper, we explore the object detection challenge with event-cameras. We propose an event-frame representation with increased temporal information that improves the performance of off the shelf feed-forward object detectors significantly. We evaluate our method on the Gen1 and 1 Mpx dataset and show that it significantly outperforms existing feed forward implementations and compare it with the current state of the art.