Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911869731602432 |
|---|---|
| author | Li, Aobo Wu, Jinjian Liu, Yongxu Li, Leida |
| author_facet | Li, Aobo Wu, Jinjian Liu, Yongxu Li, Leida |
| contents | The annotation of blind image quality assessment (BIQA) is labor-intensive and time-consuming, especially for authentic images. Training on synthetic data is expected to be beneficial, but synthetically trained models often suffer from poor generalization in real domains due to domain gaps. In this work, we make a key observation that introducing more distortion types in the synthetic dataset may not improve or even be harmful to generalizing authentic image quality assessment. To solve this challenge, we propose distortion-guided unsupervised domain adaptation for BIQA (DGQA), a novel framework that leverages adaptive multi-domain selection via prior knowledge from distortion to match the data distribution between the source domains and the target domain, thereby reducing negative transfer from the outlier source domains. Extensive experiments on two cross-domain settings (synthetic distortion to authentic distortion and synthetic distortion to algorithmic distortion) have demonstrated the effectiveness of our proposed DGQA. Besides, DGQA is orthogonal to existing model-based BIQA methods, and can be used in combination with such models to improve performance with less training data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_04167 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment Li, Aobo Wu, Jinjian Liu, Yongxu Li, Leida Computer Vision and Pattern Recognition Image and Video Processing The annotation of blind image quality assessment (BIQA) is labor-intensive and time-consuming, especially for authentic images. Training on synthetic data is expected to be beneficial, but synthetically trained models often suffer from poor generalization in real domains due to domain gaps. In this work, we make a key observation that introducing more distortion types in the synthetic dataset may not improve or even be harmful to generalizing authentic image quality assessment. To solve this challenge, we propose distortion-guided unsupervised domain adaptation for BIQA (DGQA), a novel framework that leverages adaptive multi-domain selection via prior knowledge from distortion to match the data distribution between the source domains and the target domain, thereby reducing negative transfer from the outlier source domains. Extensive experiments on two cross-domain settings (synthetic distortion to authentic distortion and synthetic distortion to algorithmic distortion) have demonstrated the effectiveness of our proposed DGQA. Besides, DGQA is orthogonal to existing model-based BIQA methods, and can be used in combination with such models to improve performance with less training data. |
| title | Bridging the Synthetic-to-Authentic Gap: Distortion-Guided Unsupervised Domain Adaptation for Blind Image Quality Assessment |
| topic | Computer Vision and Pattern Recognition Image and Video Processing |
| url | https://arxiv.org/abs/2405.04167 |