SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flow

TL;DR

SEA-RAFT is a new optical flow method that is simpler, faster, and more accurate than previous approaches, achieving state-of-the-art results with improved generalization and efficiency.

Contribution

It introduces a novel loss function, direct flow regression, and rigid-motion pre-training to enhance accuracy and speed in optical flow estimation.

Findings

Achieves 3.69 EPE on Spring benchmark, outperforming previous methods.

Operates at least 2.3x faster than existing methods.

Shows superior cross-dataset generalization on KITTI and Spring.

Abstract

We introduce SEA-RAFT, a more simple, efficient, and accurate RAFT for optical flow. Compared with RAFT, SEA-RAFT is trained with a new loss (mixture of Laplace). It directly regresses an initial flow for faster convergence in iterative refinements and introduces rigid-motion pre-training to improve generalization. SEA-RAFT achieves state-of-the-art accuracy on the Spring benchmark with a 3.69 endpoint-error (EPE) and a 0.36 1-pixel outlier rate (1px), representing 22.9% and 17.8% error reduction from best published results. In addition, SEA-RAFT obtains the best cross-dataset generalization on KITTI and Spring. With its high efficiency, SEA-RAFT operates at least 2.3x faster than existing methods while maintaining competitive performance. The code is publicly available at https://github.com/princeton-vl/SEA-RAFT.