Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2403.01582 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913996630654976 |
|---|---|
| author | Pei, Jiangbo Li, Ruizhe Men, Aidong Liu, Yang Zhuang, Xiahai Chen, Qingchao |
| author_facet | Pei, Jiangbo Li, Ruizhe Men, Aidong Liu, Yang Zhuang, Xiahai Chen, Qingchao |
| contents | Conventional Multi-Source Free Domain Adaptation (MSFDA) assumes that each source domain provides a single source model, and all source models adopt a uniform architecture. This paper introduces Zoo-MSFDA, a more general setting that allows each source domain to offer a zoo of multiple source models with different architectures. While it enriches the source knowledge, Zoo-MSFDA risks being dominated by suboptimal/harmful models. To address this issue, we theoretically analyze the model selection problem in Zoo-MSFDA, and introduce two principles: transferability principle and diversity principle. Recognizing the challenge of measuring transferability, we subsequently propose a novel Source-Free Unsupervised Transferability Estimation (SUTE). It enables assessing and comparing transferability across multiple source models with different architectures under domain shift, without requiring target labels and source data. Based on above, we introduce a Selection, Ensemble, and Adaptation (SEA) framework to address Zoo-MSFDA, which consists of: 1) source models selection based on the proposed principles and SUTE; 2) ensemble construction based on SUTE-estimated transferability; 3) target-domain adaptation of the ensemble model. Evaluations demonstrate that our SEA framework, with the introduced Zoo-MSFDA setting, significantly improves adaptation performance (e.g., 13.5% on DomainNet). Additionally, our SUTE achieves state-of-the-art performance in transferability estimation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_01582 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Selection, Ensemble, and Adaptation: Advancing Multi-Source-Free Domain Adaptation via Architecture Zoo Pei, Jiangbo Li, Ruizhe Men, Aidong Liu, Yang Zhuang, Xiahai Chen, Qingchao Machine Learning Conventional Multi-Source Free Domain Adaptation (MSFDA) assumes that each source domain provides a single source model, and all source models adopt a uniform architecture. This paper introduces Zoo-MSFDA, a more general setting that allows each source domain to offer a zoo of multiple source models with different architectures. While it enriches the source knowledge, Zoo-MSFDA risks being dominated by suboptimal/harmful models. To address this issue, we theoretically analyze the model selection problem in Zoo-MSFDA, and introduce two principles: transferability principle and diversity principle. Recognizing the challenge of measuring transferability, we subsequently propose a novel Source-Free Unsupervised Transferability Estimation (SUTE). It enables assessing and comparing transferability across multiple source models with different architectures under domain shift, without requiring target labels and source data. Based on above, we introduce a Selection, Ensemble, and Adaptation (SEA) framework to address Zoo-MSFDA, which consists of: 1) source models selection based on the proposed principles and SUTE; 2) ensemble construction based on SUTE-estimated transferability; 3) target-domain adaptation of the ensemble model. Evaluations demonstrate that our SEA framework, with the introduced Zoo-MSFDA setting, significantly improves adaptation performance (e.g., 13.5% on DomainNet). Additionally, our SUTE achieves state-of-the-art performance in transferability estimation. |
| title | Selection, Ensemble, and Adaptation: Advancing Multi-Source-Free Domain Adaptation via Architecture Zoo |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2403.01582 |