CSFlow: Learning Optical Flow via Cross Strip Correlation for Autonomous Driving

TL;DR

CSFlow introduces a novel deep network architecture with cross strip correlation and correlation regression initialization modules, significantly improving optical flow estimation accuracy for autonomous driving by effectively capturing global context.

Contribution

The paper proposes CSFlow, a new architecture with two modules that enhance global context encoding and flow initialization, advancing state-of-the-art in autonomous driving optical flow estimation.

Findings

Achieved state-of-the-art accuracy on KITTI-2015 dataset.

Effectively encodes global context with cross strip correlation.

Improves optical flow estimation in complex street scenes.

Abstract

Optical flow estimation is an essential task in self-driving systems, which helps autonomous vehicles perceive temporal continuity information of surrounding scenes. The calculation of all-pair correlation plays an important role in many existing state-of-the-art optical flow estimation methods. However, the reliance on local knowledge often limits the model's accuracy under complex street scenes. In this paper, we propose a new deep network architecture for optical flow estimation in autonomous driving--CSFlow, which consists of two novel modules: Cross Strip Correlation module (CSC) and Correlation Regression Initialization module (CRI). CSC utilizes a striping operation across the target image and the attended image to encode global context into correlation volumes, while maintaining high efficiency. CRI is used to maximally exploit the global context for optical flow initialization. Our method has achieved state-of-the-art accuracy on the public autonomous driving dataset KITTI-2015. Code is publicly available at https://github.com/MasterHow/CSFlow.