PICABench: How Far Are We from Physically Realistic Image Editing?

YuandongPu, Le Zhuo, Songhao Han, Jinbo Xing, Kaiwen Zhu, caoshuo, Bin Fu, Si Liu, Hongsheng Li, Yu Qiao, Wenlong Zhang, xichen, Yihao Liu

PICABench: How Far Are We from Physically Realistic Image Editing?: 48 upvotes on Hugging Face Daily Papers, #3 of 36 papers on 2025-10-21. Day-by-day upvote history. It lost 17 votes when the Hub removed votes in bulk.

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.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.