HALO: Human-Aligned End-to-end Image Retargeting with Layered Transformations
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908300073762816 |
|---|---|
| author | Xu, Yiran Xie, Siqi Li, Zhuofang Shadmany, Harris Li, Yinxiao Sbaiz, Luciano Wang, Miaosen Ke, Junjie Lezama, Jose Qi, Hang Zhang, Han Berent, Jesse Yang, Ming-Hsuan Essa, Irfan Huang, Jia-Bin Yang, Feng |
| author_facet | Xu, Yiran Xie, Siqi Li, Zhuofang Shadmany, Harris Li, Yinxiao Sbaiz, Luciano Wang, Miaosen Ke, Junjie Lezama, Jose Qi, Hang Zhang, Han Berent, Jesse Yang, Ming-Hsuan Essa, Irfan Huang, Jia-Bin Yang, Feng |
| contents | Image retargeting aims to change the aspect-ratio of an image while maintaining its content and structure with less visual artifacts. Existing methods still generate many artifacts or fail to maintain original content or structure. To address this, we introduce HALO, an end-to-end trainable solution for image retargeting. Since humans are more sensitive to distortions in salient areas than non-salient areas of an image, HALO decomposes the input image into salient/non-salient layers and applies different wrapping fields to different layers. To further minimize the structure distortion in the output images, we propose perceptual structure similarity loss which measures the structure similarity between input and output images and aligns with human perception. Both quantitative results and a user study on the RetargetMe dataset show that HALO achieves SOTA. Especially, our method achieves an 18.4% higher user preference compared to the baselines on average. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_03026 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | HALO: Human-Aligned End-to-end Image Retargeting with Layered Transformations Xu, Yiran Xie, Siqi Li, Zhuofang Shadmany, Harris Li, Yinxiao Sbaiz, Luciano Wang, Miaosen Ke, Junjie Lezama, Jose Qi, Hang Zhang, Han Berent, Jesse Yang, Ming-Hsuan Essa, Irfan Huang, Jia-Bin Yang, Feng Computer Vision and Pattern Recognition Image retargeting aims to change the aspect-ratio of an image while maintaining its content and structure with less visual artifacts. Existing methods still generate many artifacts or fail to maintain original content or structure. To address this, we introduce HALO, an end-to-end trainable solution for image retargeting. Since humans are more sensitive to distortions in salient areas than non-salient areas of an image, HALO decomposes the input image into salient/non-salient layers and applies different wrapping fields to different layers. To further minimize the structure distortion in the output images, we propose perceptual structure similarity loss which measures the structure similarity between input and output images and aligns with human perception. Both quantitative results and a user study on the RetargetMe dataset show that HALO achieves SOTA. Especially, our method achieves an 18.4% higher user preference compared to the baselines on average. |
| title | HALO: Human-Aligned End-to-end Image Retargeting with Layered Transformations |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.03026 |