Fine-Grained Human Pose Editing Assessment via Layer-Selective MLLMs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912831532695552 |
|---|---|
| author | Sun, Ningyu Cai, Zhaolin Xu, Zitong Chen, Peihang Duan, Huiyu Yan, Yichao Min, Xiongkuo Yang, Xiaokang |
| author_facet | Sun, Ningyu Cai, Zhaolin Xu, Zitong Chen, Peihang Duan, Huiyu Yan, Yichao Min, Xiongkuo Yang, Xiaokang |
| contents | Text-guided human pose editing has gained significant traction in AIGC applications. However,it remains plagued by structural anomalies and generative artifacts. Existing evaluation metrics often isolate authenticity detection from quality assessment, failing to provide fine-grained insights into pose-specific inconsistencies. To address these limitations, we introduce HPE-Bench, a specialized benchmark comprising 1,700 standardized samples from 17 state-of-the-art editing models, offering both authenticity labels and multi-dimensional quality scores. Furthermore, we propose a unified framework based on layer-selective multimodal large language models (MLLMs). By employing contrastive LoRA tuning and a novel layer sensitivity analysis (LSA) mechanism, we identify the optimal feature layer for pose evaluation. Our framework achieves superior performance in both authenticity detection and multi-dimensional quality regression, effectively bridging the gap between forensic detection and quality assessment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_10369 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Fine-Grained Human Pose Editing Assessment via Layer-Selective MLLMs Sun, Ningyu Cai, Zhaolin Xu, Zitong Chen, Peihang Duan, Huiyu Yan, Yichao Min, Xiongkuo Yang, Xiaokang Computer Vision and Pattern Recognition Text-guided human pose editing has gained significant traction in AIGC applications. However,it remains plagued by structural anomalies and generative artifacts. Existing evaluation metrics often isolate authenticity detection from quality assessment, failing to provide fine-grained insights into pose-specific inconsistencies. To address these limitations, we introduce HPE-Bench, a specialized benchmark comprising 1,700 standardized samples from 17 state-of-the-art editing models, offering both authenticity labels and multi-dimensional quality scores. Furthermore, we propose a unified framework based on layer-selective multimodal large language models (MLLMs). By employing contrastive LoRA tuning and a novel layer sensitivity analysis (LSA) mechanism, we identify the optimal feature layer for pose evaluation. Our framework achieves superior performance in both authenticity detection and multi-dimensional quality regression, effectively bridging the gap between forensic detection and quality assessment. |
| title | Fine-Grained Human Pose Editing Assessment via Layer-Selective MLLMs |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.10369 |