Rethinking the Role of Pre-Trained Networks in Source-Free Domain Adaptation
- Resource Type
- Conference
- Authors
- Zhang, Wenyu; Shen, Li; Foo, Chuan-Sheng
- Source
- 2023 IEEE/CVF International Conference on Computer Vision (ICCV) ICCV Computer Vision (ICCV), 2023 IEEE/CVF International Conference on. :18795-18805 Oct, 2023
- Subject
- Computing and Processing
Signal Processing and Analysis
Training
Representation learning
Adaptation models
Computer vision
Computational modeling
Benchmark testing
Robustness
- Language
- ISSN
- 2380-7504
Source-free domain adaptation (SFDA) aims to adapt a source model trained on a fully-labeled source domain to an unlabeled target domain. Large-data pre-trained networks are used to initialize source models during source training, and subsequently discarded. However, source training can cause the model to overfit to source data distribution and lose applicable target domain knowledge. We propose to integrate the pre-trained network into the target adaptation process as it has diversified features important for generalization and provides an alternate view of features and classification decisions different from the source model. We propose to distil useful target domain information through a co-learning strategy to improve target pseudolabel quality for finetuning the source model. Evaluation on 4 benchmark datasets show that our proposed strategy improves adaptation performance and can be successfully integrated with existing SFDA methods. Leveraging modern pre-trained networks that have stronger representation learning ability in the co-learning strategy further boosts performance.