Source-Free Domain Adaptation for Semantic Segmentation

By Yuang Liu, Wei Zhang, Jun Wang. Computer Vision and Pattern Recognition, 2021

332 citations recorded by the source corpus.

Unsupervised Domain Adaptation (UDA) can tackle the challenge that convolutional neural network (CNN)-based approaches for semantic segmentation heavily rely on the pixel-level annotated data, which is labor-intensive. However, existing UDA approaches in this regard inevitably require the full access to source datasets to reduce the gap between the source and target domains during model adaptation, which are impractical in the real scenarios where the source datasets are private, and thus cannot be released along with the well-trained source models. To cope with this issue, we propose a source-free domain adaptation framework for semantic segmentation, namely SFDA, in which only a well-trained source model and an unlabeled target domain dataset are available for adaptation. SFDA not only enables to recover and preserve the source domain knowledge from the source model via knowledge transfer during model adaptation, but also distills valuable information from the target domain for self-supervised learning. The pixel-and patch-level optimization objectives tailored for semantic segmentation are seamlessly integrated in the framework. The extensive experimental results on numerous benchmark datasets highlight the effectiveness of our framework against the existing UDA approaches relying on source data.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/a633ea2dea6a9bd24942e1f7c0cea4f7a94670df

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Joost van de Weijer
2. Jian Liang
3. Dengxin Dai
4. Luc Van Gool
5. Luis Herranz
6. Ran He
7. Zhaoxiang Zhang
8. Yuxi Wang

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Joost van de Weijer
3. Dengxin Dai
4. Ran He
5. Dacheng Tao
6. Jian Liang
7. Ming-Hsuan Yang
8. Tien-Ping Tan

In Radiology, Nuclear Medicine and Imaging

1. Luc Van Gool
2. Dengxin Dai
3. Zhaoxiang Zhang
4. Yuxi Wang
5. Luis Herranz
6. Ran He
7. Tien-Ping Tan
8. Jian Liang

Citations in the corpus

• Exploiting the Intrinsic Neighborhood Structure for Source-free Domain Adaptation (2021, methodology): “We exploit this intrinsic structure of the target data for SFDA by considering the neighborhood information, and the adaptation is achieved with the following objective: L = − 1 nt ∑ xi∈Dt ∑ xj∈Neigh(xi) Dsim(pi, pj) Ddis(xi, xj) (1) where the Neigh(xi) means the nearest neighbors of xi, Dsim comput”
• Domain Adaptation for Medical Image Analysis: A Survey (2021, neutral): “Source-free domain adaptation, an emerging topic that avoids potential violation of data privacy policies, has drawn increasing attention recently [155]–[158].”
• Continual Test-Time Domain Adaptation (2022, background): “While most existing works focus on image classification, [20,27,38] extend test-time adaptation to semantic segmentation.”
• A Comprehensive Survey on Test-Time Adaptation Under Distribution Shifts (2023, positive): “Motivated by recent advances in data-free knowledge distillation [23], [82], SFDA-KTMA [284] exploits the moving average statistics of activations stored in BN layers of the pre-trained source model and imposes the following BN”
• Give Me Your Trained Model: Domain Adaptive Semantic Segmentation without Source Data (2021, methodology): “Recent studies employ model adaptation [19] or generated synthetic fake source data [20] to tackle the source-data absent problem.”