Unbiased Video Scene Graph Generation via Visual and Semantic Dual Debiasing
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_ | 1866908302353367040 |
|---|---|
| author | Li, Yanjun Li, Zhaoyang Chen, Honghui Xu, Lizhi |
| author_facet | Li, Yanjun Li, Zhaoyang Chen, Honghui Xu, Lizhi |
| contents | Video Scene Graph Generation (VidSGG) aims to capture dynamic relationships among entities by sequentially analyzing video frames and integrating visual and semantic information. However, VidSGG is challenged by significant biases that skew predictions. To mitigate these biases, we propose a VIsual and Semantic Awareness (VISA) framework for unbiased VidSGG. VISA addresses visual bias through memory-enhanced temporal integration that enhances object representations and concurrently reduces semantic bias by iteratively integrating object features with comprehensive semantic information derived from triplet relationships. This visual-semantics dual debiasing approach results in more unbiased representations of complex scene dynamics. Extensive experiments demonstrate the effectiveness of our method, where VISA outperforms existing unbiased VidSGG approaches by a substantial margin (e.g., +13.1% improvement in mR@20 and mR@50 for the SGCLS task under Semi Constraint). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_00548 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Unbiased Video Scene Graph Generation via Visual and Semantic Dual Debiasing Li, Yanjun Li, Zhaoyang Chen, Honghui Xu, Lizhi Computer Vision and Pattern Recognition Multimedia Video Scene Graph Generation (VidSGG) aims to capture dynamic relationships among entities by sequentially analyzing video frames and integrating visual and semantic information. However, VidSGG is challenged by significant biases that skew predictions. To mitigate these biases, we propose a VIsual and Semantic Awareness (VISA) framework for unbiased VidSGG. VISA addresses visual bias through memory-enhanced temporal integration that enhances object representations and concurrently reduces semantic bias by iteratively integrating object features with comprehensive semantic information derived from triplet relationships. This visual-semantics dual debiasing approach results in more unbiased representations of complex scene dynamics. Extensive experiments demonstrate the effectiveness of our method, where VISA outperforms existing unbiased VidSGG approaches by a substantial margin (e.g., +13.1% improvement in mR@20 and mR@50 for the SGCLS task under Semi Constraint). |
| title | Unbiased Video Scene Graph Generation via Visual and Semantic Dual Debiasing |
| topic | Computer Vision and Pattern Recognition Multimedia |
| url | https://arxiv.org/abs/2503.00548 |