FAME: Fairness-aware Attention-modulated Video 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_ | 1866911234225340416 |
|---|---|
| author | Wu, Zhangkai Fan, Xuhui Xie, Zhongyuan Shi, Kaize Li, Zhidong Cao, Longbing |
| author_facet | Wu, Zhangkai Fan, Xuhui Xie, Zhongyuan Shi, Kaize Li, Zhidong Cao, Longbing |
| contents | Training-free video editing (VE) models tend to fall back on gender stereotypes when rendering profession-related prompts. We propose \textbf{FAME} for \textit{Fairness-aware Attention-modulated Video Editing} that mitigates profession-related gender biases while preserving prompt alignment and temporal consistency for coherent VE. We derive fairness embeddings from existing minority representations by softly injecting debiasing tokens into the text encoder. Simultaneously, FAME integrates fairness modulation into both temporal self attention and prompt-to-region cross attention to mitigate the motion corruption and temporal inconsistency caused by directly introducing fairness cues. For temporal self attention, FAME introduces a region constrained attention mask combined with time decay weighting, which enhances intra-region coherence while suppressing irrelevant inter-region interactions. For cross attention, it reweights tokens to region matching scores by incorporating fairness sensitive similarity masks derived from debiasing prompt embeddings. Together, these modulations keep fairness-sensitive semantics tied to the right visual regions and prevent temporal drift across frames. Extensive experiments on new VE fairness-oriented benchmark \textit{FairVE} demonstrate that FAME achieves stronger fairness alignment and semantic fidelity, surpassing existing VE baselines. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_22960 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FAME: Fairness-aware Attention-modulated Video Editing Wu, Zhangkai Fan, Xuhui Xie, Zhongyuan Shi, Kaize Li, Zhidong Cao, Longbing Computer Vision and Pattern Recognition Artificial Intelligence Training-free video editing (VE) models tend to fall back on gender stereotypes when rendering profession-related prompts. We propose \textbf{FAME} for \textit{Fairness-aware Attention-modulated Video Editing} that mitigates profession-related gender biases while preserving prompt alignment and temporal consistency for coherent VE. We derive fairness embeddings from existing minority representations by softly injecting debiasing tokens into the text encoder. Simultaneously, FAME integrates fairness modulation into both temporal self attention and prompt-to-region cross attention to mitigate the motion corruption and temporal inconsistency caused by directly introducing fairness cues. For temporal self attention, FAME introduces a region constrained attention mask combined with time decay weighting, which enhances intra-region coherence while suppressing irrelevant inter-region interactions. For cross attention, it reweights tokens to region matching scores by incorporating fairness sensitive similarity masks derived from debiasing prompt embeddings. Together, these modulations keep fairness-sensitive semantics tied to the right visual regions and prevent temporal drift across frames. Extensive experiments on new VE fairness-oriented benchmark \textit{FairVE} demonstrate that FAME achieves stronger fairness alignment and semantic fidelity, surpassing existing VE baselines. |
| title | FAME: Fairness-aware Attention-modulated Video Editing |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2510.22960 |