UniSER: A Foundation Model for Unified Soft Effects Removal
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_ | 1866913066470342656 |
|---|---|
| author | Zhang, Jingdong Zhang, Lingzhi Liu, Qing Chiu, Mang Tik Barnes, Connelly Wang, Yizhou You, Haoran Liu, Xiaoyang Zhou, Yuqian Lin, Zhe Shechtman, Eli Amirghodsi, Sohrab Li, Xin Wang, Wenping Zhan, Xiaohang |
| author_facet | Zhang, Jingdong Zhang, Lingzhi Liu, Qing Chiu, Mang Tik Barnes, Connelly Wang, Yizhou You, Haoran Liu, Xiaoyang Zhou, Yuqian Lin, Zhe Shechtman, Eli Amirghodsi, Sohrab Li, Xin Wang, Wenping Zhan, Xiaohang |
| contents | Digital images are often degraded by soft effects such as lens flare, haze, shadows, and reflections, which reduce aesthetics even though the underlying pixels remain partially visible. The prevailing works address these degradations in isolation, developing highly specialized, specialist models that lack scalability and fail to exploit the shared underlying essences of these restoration problems. Meanwhile, although recent large-scale generalist models (e.g., GPT-4o, Flux Kontext, Nano Banana) offer powerful text-driven editing capabilities, they heavily rely on detailed prompts and often fail to achieve robust removal on such fine-grained tasks while preserving the scene's identity. Leveraging the common essence of soft effects, i.e., semi-transparent occlusions, we introduce a foundational versatile model UniSER, capable of addressing diverse degradations caused by soft effects within a single framework. Our methodology centers on curating a massive 3.8M-pair dataset to ensure robustness and generalization, which includes novel, physically-plausible data to fill critical gaps in public benchmarks, and a tailored training pipeline that fine-tunes a Diffusion Transformer to learn robust restoration priors from this diverse data, integrating fine-grained mask and strength controls. This synergistic approach allows UniSER to significantly outperform both specialist and generalist models, achieving robust, high-fidelity restoration in the wild. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_14183 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UniSER: A Foundation Model for Unified Soft Effects Removal Zhang, Jingdong Zhang, Lingzhi Liu, Qing Chiu, Mang Tik Barnes, Connelly Wang, Yizhou You, Haoran Liu, Xiaoyang Zhou, Yuqian Lin, Zhe Shechtman, Eli Amirghodsi, Sohrab Li, Xin Wang, Wenping Zhan, Xiaohang Computer Vision and Pattern Recognition Digital images are often degraded by soft effects such as lens flare, haze, shadows, and reflections, which reduce aesthetics even though the underlying pixels remain partially visible. The prevailing works address these degradations in isolation, developing highly specialized, specialist models that lack scalability and fail to exploit the shared underlying essences of these restoration problems. Meanwhile, although recent large-scale generalist models (e.g., GPT-4o, Flux Kontext, Nano Banana) offer powerful text-driven editing capabilities, they heavily rely on detailed prompts and often fail to achieve robust removal on such fine-grained tasks while preserving the scene's identity. Leveraging the common essence of soft effects, i.e., semi-transparent occlusions, we introduce a foundational versatile model UniSER, capable of addressing diverse degradations caused by soft effects within a single framework. Our methodology centers on curating a massive 3.8M-pair dataset to ensure robustness and generalization, which includes novel, physically-plausible data to fill critical gaps in public benchmarks, and a tailored training pipeline that fine-tunes a Diffusion Transformer to learn robust restoration priors from this diverse data, integrating fine-grained mask and strength controls. This synergistic approach allows UniSER to significantly outperform both specialist and generalist models, achieving robust, high-fidelity restoration in the wild. |
| title | UniSER: A Foundation Model for Unified Soft Effects Removal |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2511.14183 |