Multi-Target Adversarial Frameworks for Domain Adaptation in Semantic Segmentation

TL;DR

This paper introduces two adversarial frameworks for unsupervised domain adaptation in semantic segmentation, enabling a single model to handle multiple target domains effectively, which is crucial for real-world autonomous systems.

Contribution

It proposes novel multi-target adversarial frameworks, including multi-discriminator and multi-target knowledge transfer, for improved domain adaptation in semantic segmentation.

Findings

Outperforms baseline methods on four new multi-target benchmarks

Consistently improves segmentation accuracy across multiple target domains

Sets new standards for multi-target domain adaptation in semantic segmentation

Abstract

In this work, we address the task of unsupervised domain adaptation (UDA) for semantic segmentation in presence of multiple target domains: The objective is to train a single model that can handle all these domains at test time. Such a multi-target adaptation is crucial for a variety of scenarios that real-world autonomous systems must handle. It is a challenging setup since one faces not only the domain gap between the labeled source set and the unlabeled target set, but also the distribution shifts existing within the latter among the different target domains. To this end, we introduce two adversarial frameworks: (i) multi-discriminator, which explicitly aligns each target domain to its counterparts, and (ii) multi-target knowledge transfer, which learns a target-agnostic model thanks to a multi-teacher/single-student distillation mechanism.The evaluation is done on four newly-proposed multi-target benchmarks for UDA in semantic segmentation. In all tested scenarios, our approaches consistently outperform baselines, setting competitive standards for the novel task.