EventSSEG: Event-driven Self-Supervised Segmentation with Probabilistic Attention

TL;DR

EventSSEG introduces an event-driven, self-supervised road segmentation method that leverages probabilistic attention and requires minimal labeled data, achieving state-of-the-art results with event cameras.

Contribution

It presents a novel self-supervised learning approach for event camera-based segmentation, overcoming the challenge of transferring pretrained weights and reducing labeled data dependency.

Findings

Achieves state-of-the-art performance on DSEC-Semantic and DDD17 datasets.

Operates with minimal labeled events, demonstrating efficiency.

Utilizes event-only computing with probabilistic attention for low-latency segmentation.

Abstract

Road segmentation is pivotal for autonomous vehicles, yet achieving low latency and low compute solutions using frame based cameras remains a challenge. Event cameras offer a promising alternative. To leverage their low power sensing, we introduce EventSSEG, a method for road segmentation that uses event only computing and a probabilistic attention mechanism. Event only computing poses a challenge in transferring pretrained weights from the conventional camera domain, requiring abundant labeled data, which is scarce. To overcome this, EventSSEG employs event-based self supervised learning, eliminating the need for extensive labeled data. Experiments on DSEC-Semantic and DDD17 show that EventSSEG achieves state of the art performance with minimal labeled events. This approach maximizes event cameras capabilities and addresses the lack of labeled events.