Epipolar Attention Field Transformers for Bird's Eye View Semantic Segmentation

TL;DR

This paper introduces Epipolar Attention Fields in transformer models to improve bird's eye view semantic segmentation for autonomous driving, reducing reliance on learned positional encodings and enhancing generalization.

Contribution

It proposes a novel epipolar geometric constraint-based attention mechanism, EAFormer, that outperforms previous methods and improves generalization in BEV semantic segmentation.

Findings

EAFormer achieves 2% higher mIoU in map segmentation.

The method demonstrates better generalization to unseen camera configurations.

Epipolar Attention Fields effectively replace learned positional encodings.

Abstract

Spatial understanding of the semantics of the surroundings is a key capability needed by autonomous cars to enable safe driving decisions. Recently, purely vision-based solutions have gained increasing research interest. In particular, approaches extracting a bird's eye view (BEV) from multiple cameras have demonstrated great performance for spatial understanding. This paper addresses the dependency on learned positional encodings to correlate image and BEV feature map elements for transformer-based methods. We propose leveraging epipolar geometric constraints to model the relationship between cameras and the BEV by Epipolar Attention Fields. They are incorporated into the attention mechanism as a novel attribution term, serving as an alternative to learned positional encodings. Experiments show that our method EAFormer outperforms previous BEV approaches by 2% mIoU for map semantic segmentation and exhibits superior generalization capabilities compared to implicitly learning the camera configuration.