E-RAFT: Dense Optical Flow from Event Cameras

TL;DR

This paper introduces E-RAFT, a novel dense optical flow method for event cameras that incorporates feature correlation and sequential processing, significantly improving accuracy and introducing a new high-resolution dataset.

Contribution

It presents a new dense optical flow approach for event cameras that leverages feature correlation, outperforming previous methods and providing a new dataset with larger displacements.

Findings

Reduces end-point error by 23% on MVSEC dataset.

Achieves 66% error reduction on a new high-resolution dataset.

Introduces a novel dataset with larger displacements for event camera optical flow.

Abstract

We propose to incorporate feature correlation and sequential processing into dense optical flow estimation from event cameras. Modern frame-based optical flow methods heavily rely on matching costs computed from feature correlation. In contrast, there exists no optical flow method for event cameras that explicitly computes matching costs. Instead, learning-based approaches using events usually resort to the U-Net architecture to estimate optical flow sparsely. Our key finding is that the introduction of correlation features significantly improves results compared to previous methods that solely rely on convolution layers. Compared to the state-of-the-art, our proposed approach computes dense optical flow and reduces the end-point error by 23% on MVSEC. Furthermore, we show that all existing optical flow methods developed so far for event cameras have been evaluated on datasets with very small displacement fields with a maximum flow magnitude of 10 pixels. Based on this observation, we introduce a new real-world dataset that exhibits displacement fields with magnitudes up to 210 pixels and 3 times higher camera resolution. Our proposed approach reduces the end-point error on this dataset by 66%.