REFINE-CONTROL: A Semi-supervised Distillation Method For Conditional Image Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909808930586624 |
|---|---|
| author | Jiang, Yicheng Yuan, Jin Yuan, Hua Zhang, Yao Rui, Yong |
| author_facet | Jiang, Yicheng Yuan, Jin Yuan, Hua Zhang, Yao Rui, Yong |
| contents | Conditional image generation models have achieved remarkable results by leveraging text-based control to generate customized images. However, the high resource demands of these models and the scarcity of well-annotated data have hindered their deployment on edge devices, leading to enormous costs and privacy concerns, especially when user data is sent to a third party. To overcome these challenges, we propose Refine-Control, a semi-supervised distillation framework. Specifically, we improve the performance of the student model by introducing a tri-level knowledge fusion loss to transfer different levels of knowledge. To enhance generalization and alleviate dataset scarcity, we introduce a semi-supervised distillation method utilizing both labeled and unlabeled data. Our experiments reveal that Refine-Control achieves significant reductions in computational cost and latency, while maintaining high-fidelity generation capabilities and controllability, as quantified by comparative metrics. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_22139 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | REFINE-CONTROL: A Semi-supervised Distillation Method For Conditional Image Generation Jiang, Yicheng Yuan, Jin Yuan, Hua Zhang, Yao Rui, Yong Computer Vision and Pattern Recognition Artificial Intelligence Conditional image generation models have achieved remarkable results by leveraging text-based control to generate customized images. However, the high resource demands of these models and the scarcity of well-annotated data have hindered their deployment on edge devices, leading to enormous costs and privacy concerns, especially when user data is sent to a third party. To overcome these challenges, we propose Refine-Control, a semi-supervised distillation framework. Specifically, we improve the performance of the student model by introducing a tri-level knowledge fusion loss to transfer different levels of knowledge. To enhance generalization and alleviate dataset scarcity, we introduce a semi-supervised distillation method utilizing both labeled and unlabeled data. Our experiments reveal that Refine-Control achieves significant reductions in computational cost and latency, while maintaining high-fidelity generation capabilities and controllability, as quantified by comparative metrics. |
| title | REFINE-CONTROL: A Semi-supervised Distillation Method For Conditional Image Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2509.22139 |