GFRRN: Explore the Gaps in Single Image Reflection Removal
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908855745642496 |
|---|---|
| author | Chen, Yu He, Zewei Liu, Xingyu Chen, Zixuan Lu, Zheming |
| author_facet | Chen, Yu He, Zewei Liu, Xingyu Chen, Zixuan Lu, Zheming |
| contents | Prior dual-stream methods with the feature interaction mechanism have achieved remarkable performance in single image reflection removal (SIRR). However, they often struggle with (1) semantic understanding gap between the features of pre-trained models and those of reflection removal models, and (2) reflection label inconsistencies between synthetic and real-world training data. In this work, we first adopt the parameter efficient fine-tuning (PEFT) strategy by integrating several learnable Mona layers into the pre-trained model to align the training directions. Then, a label generator is designed to unify the reflection labels for both synthetic and real-world data. In addition, a Gaussian-based Adaptive Frequency Learning Block (G-AFLB) is proposed to adaptively learn and fuse the frequency priors, and a Dynamic Agent Attention (DAA) is employed as an alternative to window-based attention by dynamically modeling the significance levels across windows (inter-) and within an individual window (intra-). These components constitute our proposed Gap-Free Reflection Removal Network (GFRRN). Extensive experiments demonstrate the effectiveness of our GFRRN, achieving superior performance against state-of-the-art SIRR methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_22695 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | GFRRN: Explore the Gaps in Single Image Reflection Removal Chen, Yu He, Zewei Liu, Xingyu Chen, Zixuan Lu, Zheming Computer Vision and Pattern Recognition Prior dual-stream methods with the feature interaction mechanism have achieved remarkable performance in single image reflection removal (SIRR). However, they often struggle with (1) semantic understanding gap between the features of pre-trained models and those of reflection removal models, and (2) reflection label inconsistencies between synthetic and real-world training data. In this work, we first adopt the parameter efficient fine-tuning (PEFT) strategy by integrating several learnable Mona layers into the pre-trained model to align the training directions. Then, a label generator is designed to unify the reflection labels for both synthetic and real-world data. In addition, a Gaussian-based Adaptive Frequency Learning Block (G-AFLB) is proposed to adaptively learn and fuse the frequency priors, and a Dynamic Agent Attention (DAA) is employed as an alternative to window-based attention by dynamically modeling the significance levels across windows (inter-) and within an individual window (intra-). These components constitute our proposed Gap-Free Reflection Removal Network (GFRRN). Extensive experiments demonstrate the effectiveness of our GFRRN, achieving superior performance against state-of-the-art SIRR methods. |
| title | GFRRN: Explore the Gaps in Single Image Reflection Removal |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.22695 |