Solution for CVPR 2024 UG2+ Challenge Track on All Weather Semantic Segmentation

TL;DR

This paper presents a robust semantic segmentation solution for adverse weather conditions in the CVPR 2024 UG2+ Challenge, utilizing a pre-trained backbone, data augmentation, and state-of-the-art methods to achieve top-tier results.

Contribution

The authors introduce a novel combination of pre-trained InternImage-H backbone, data augmentation, and Upernet segmentation to improve performance in adverse weather semantic segmentation.

Findings

Achieved 3rd place in the CVPR 2024 UG2+ Challenge

Enhanced segmentation accuracy with data augmentation techniques

Demonstrated robustness across various weather conditions

Abstract

In this report, we present our solution for the semantic segmentation in adverse weather, in UG2+ Challenge at CVPR 2024. To achieve robust and accurate segmentation results across various weather conditions, we initialize the InternImage-H backbone with pre-trained weights from the large-scale joint dataset and enhance it with the state-of-the-art Upernet segmentation method. Specifically, we utilize offline and online data augmentation approaches to extend the train set, which helps us to further improve the performance of the segmenter. As a result, our proposed solution demonstrates advanced performance on the test set and achieves 3rd position in this challenge.