PICABench: How Far Are We from Physically Realistic Image Editing?
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_ | 1866918270489067520 |
|---|---|
| author | Pu, Yuandong Zhuo, Le Han, Songhao Xing, Jinbo Zhu, Kaiwen Cao, Shuo Fu, Bin Liu, Si Li, Hongsheng Qiao, Yu Zhang, Wenlong Chen, Xi Liu, Yihao |
| author_facet | Pu, Yuandong Zhuo, Le Han, Songhao Xing, Jinbo Zhu, Kaiwen Cao, Shuo Fu, Bin Liu, Si Li, Hongsheng Qiao, Yu Zhang, Wenlong Chen, Xi Liu, Yihao |
| contents | Image editing has achieved remarkable progress recently. Modern editing models could already follow complex instructions to manipulate the original content. However, beyond completing the editing instructions, the accompanying physical effects are the key to the generation realism. For example, removing an object should also remove its shadow, reflections, and interactions with nearby objects. Unfortunately, existing models and benchmarks mainly focus on instruction completion but overlook these physical effects. So, at this moment, how far are we from physically realistic image editing? To answer this, we introduce PICABench, which systematically evaluates physical realism across eight sub-dimension (spanning optics, mechanics, and state transitions) for most of the common editing operations (add, remove, attribute change, etc.). We further propose the PICAEval, a reliable evaluation protocol that uses VLM-as-a-judge with per-case, region-level human annotations and questions. Beyond benchmarking, we also explore effective solutions by learning physics from videos and construct a training dataset PICA-100K. After evaluating most of the mainstream models, we observe that physical realism remains a challenging problem with large rooms to explore. We hope that our benchmark and proposed solutions can serve as a foundation for future work moving from naive content editing toward physically consistent realism. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_17681 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PICABench: How Far Are We from Physically Realistic Image Editing? Pu, Yuandong Zhuo, Le Han, Songhao Xing, Jinbo Zhu, Kaiwen Cao, Shuo Fu, Bin Liu, Si Li, Hongsheng Qiao, Yu Zhang, Wenlong Chen, Xi Liu, Yihao Computer Vision and Pattern Recognition Artificial Intelligence Image editing has achieved remarkable progress recently. Modern editing models could already follow complex instructions to manipulate the original content. However, beyond completing the editing instructions, the accompanying physical effects are the key to the generation realism. For example, removing an object should also remove its shadow, reflections, and interactions with nearby objects. Unfortunately, existing models and benchmarks mainly focus on instruction completion but overlook these physical effects. So, at this moment, how far are we from physically realistic image editing? To answer this, we introduce PICABench, which systematically evaluates physical realism across eight sub-dimension (spanning optics, mechanics, and state transitions) for most of the common editing operations (add, remove, attribute change, etc.). We further propose the PICAEval, a reliable evaluation protocol that uses VLM-as-a-judge with per-case, region-level human annotations and questions. Beyond benchmarking, we also explore effective solutions by learning physics from videos and construct a training dataset PICA-100K. After evaluating most of the mainstream models, we observe that physical realism remains a challenging problem with large rooms to explore. We hope that our benchmark and proposed solutions can serve as a foundation for future work moving from naive content editing toward physically consistent realism. |
| title | PICABench: How Far Are We from Physically Realistic Image Editing? |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2510.17681 |