HGFormer: Hierarchical Grouping Transformer for Domain Generalized Semantic Segmentation

TL;DR

This paper introduces HGFormer, a hierarchical grouping transformer that explicitly segments parts and wholes at multiple scales to improve domain generalization in semantic segmentation tasks, outperforming existing methods.

Contribution

The paper proposes a novel hierarchical grouping transformer (HGFormer) that explicitly groups pixels at multiple scales for improved domain generalization in semantic segmentation.

Findings

HGFormer outperforms previous methods significantly in cross-domain segmentation tasks.

Explicit multi-scale grouping improves robustness over per-pixel classification.

HGFormer demonstrates superior performance across seven public datasets.

Abstract

Current semantic segmentation models have achieved great success under the independent and identically distributed (i.i.d.) condition. However, in real-world applications, test data might come from a different domain than training data. Therefore, it is important to improve model robustness against domain differences. This work studies semantic segmentation under the domain generalization setting, where a model is trained only on the source domain and tested on the unseen target domain. Existing works show that Vision Transformers are more robust than CNNs and show that this is related to the visual grouping property of self-attention. In this work, we propose a novel hierarchical grouping transformer (HGFormer) to explicitly group pixels to form part-level masks and then whole-level masks. The masks at different scales aim to segment out both parts and a whole of classes. HGFormer combines mask classification results at both scales for class label prediction. We assemble multiple interesting cross-domain settings by using seven public semantic segmentation datasets. Experiments show that HGFormer yields more robust semantic segmentation results than per-pixel classification methods and flat grouping transformers, and outperforms previous methods significantly. Code will be available at https://github.com/dingjiansw101/HGFormer.