Multi-Target Unsupervised Domain Adaptation for Semantic Segmentation without External Data

TL;DR

This paper introduces a scalable multi-target unsupervised domain adaptation method for semantic segmentation that adapts to new target domains without requiring access to external data, outperforming existing solutions.

Contribution

It proposes a novel strategy for domain adaptation that does not rely on external data during adaptation, enhancing scalability and applicability to unseen domains.

Findings

Significantly outperforms state-of-the-art methods on benchmark datasets.

Effective in synthetic-to-real and real-to-real adaptation scenarios.

Maintains knowledge from external data during adaptation without retraining.

Abstract

Multi-target unsupervised domain adaptation (UDA) aims to learn a unified model to address the domain shift between multiple target domains. Due to the difficulty of obtaining annotations for dense predictions, it has recently been introduced into cross-domain semantic segmentation. However, most existing solutions require labeled data from the source domain and unlabeled data from multiple target domains concurrently during training. Collectively, we refer to this data as "external". When faced with new unlabeled data from an unseen target domain, these solutions either do not generalize well or require retraining from scratch on all data. To address these challenges, we introduce a new strategy called "multi-target UDA without external data" for semantic segmentation. Specifically, the segmentation model is initially trained on the external data. Then, it is adapted to a new unseen target domain without accessing any external data. This approach is thus more scalable than existing solutions and remains applicable when external data is inaccessible. We demonstrate this strategy using a simple method that incorporates self-distillation and adversarial learning, where knowledge acquired from the external data is preserved during adaptation through "one-way" adversarial learning. Extensive experiments in several synthetic-to-real and real-to-real adaptation settings on four benchmark urban driving datasets show that our method significantly outperforms current state-of-the-art solutions, even in the absence of external data. Our source code is available online (https://github.com/YonghaoXu/UT-KD).