CoMaL: Conditional Maximum Likelihood Approach to Self-supervised Domain Adaptation in Long-tail Semantic Segmentation

TL;DR

This paper introduces CoMaL, a novel self-supervised domain adaptation method for long-tail semantic segmentation that models structural dependencies and outperforms existing approaches on major benchmarks.

Contribution

The paper proposes a new metric and the Conditional Maximum Likelihood (CoMaL) approach within an autoregressive framework for better long-tail domain adaptation in segmentation.

Findings

Outperforms prior methods on SYNTHIA to Cityscapes and GTA to Cityscapes benchmarks.

Effectively models structural dependencies in long-tail pixel distributions.

Achieves significant improvements in both standard and proposed evaluation protocols.

Abstract

The research in self-supervised domain adaptation in semantic segmentation has recently received considerable attention. Although GAN-based methods have become one of the most popular approaches to domain adaptation, they have suffered from some limitations. They are insufficient to model both global and local structures of a given image, especially in small regions of tail classes. Moreover, they perform bad on the tail classes containing limited number of pixels or less training samples. In order to address these issues, we present a new self-supervised domain adaptation approach to tackle long-tail semantic segmentation in this paper. Firstly, a new metric is introduced to formulate long-tail domain adaptation in the segmentation problem. Secondly, a new Conditional Maximum Likelihood (CoMaL) approach in an autoregressive framework is presented to solve the problem of long-tail domain adaptation. Although other segmentation methods work under the pixel independence assumption, the long-tailed pixel distributions in CoMaL are generally solved in the context of structural dependency, as that is more realistic. Finally, the proposed method is evaluated on popular large-scale semantic segmentation benchmarks, i.e., "SYNTHIA to Cityscapes" and "GTA to Cityscapes", and outperforms the prior methods by a large margin in both the standard and the proposed evaluation protocols.