EDeNN: Event Decay Neural Networks for low latency vision

TL;DR

EDeNN introduces event decay neural networks that process asynchronous event camera data directly, achieving low latency and state-of-the-art results in velocity and optical flow estimation without complex training.

Contribution

The paper presents a novel neural network architecture that operates directly on event streams, reducing latency and improving performance over traditional frame-based methods.

Findings

Achieves less than 1/10 the latency of existing methods

State-of-the-art angular velocity regression results

Competitive optical flow estimation performance

Abstract

Despite the success of neural networks in computer vision tasks, digital 'neurons' are a very loose approximation of biological neurons. Today's learning approaches are designed to function on digital devices with digital data representations such as image frames. In contrast, biological vision systems are generally much more capable and efficient than state-of-the-art digital computer vision algorithms. Event cameras are an emerging sensor technology which imitates biological vision with asynchronously firing pixels, eschewing the concept of the image frame. To leverage modern learning techniques, many event-based algorithms are forced to accumulate events back to image frames, somewhat squandering the advantages of event cameras. We follow the opposite paradigm and develop a new type of neural network which operates closer to the original event data stream. We demonstrate state-of-the-art performance in angular velocity regression and competitive optical flow estimation, while avoiding difficulties related to training SNN. Furthermore, the processing latency of our proposed approach is less than 1/10 any other implementation, while continuous inference increases this improvement by another order of magnitude.