Epipolar Attention Field Transformers for Bird's Eye View Semantic Segmentation
This work addresses a key problem for autonomous cars by enhancing spatial understanding through better camera-to-BEV correlation, though it is incremental as it builds on existing transformer-based methods.
The paper tackled the dependency on learned positional encodings in transformer-based methods for bird's eye view semantic segmentation by proposing Epipolar Attention Fields that incorporate epipolar geometric constraints into the attention mechanism, resulting in a 2% mIoU improvement over previous approaches.
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.