PAEFF: Precise Alignment and Enhanced Gated Feature Fusion for Face-Voice Association
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918037107507200 |
|---|---|
| author | Hannan, Abdul Manzoor, Muhammad Arslan Nawaz, Shah Liaqat, Muhammad Irzam Schedl, Markus Noman, Mubashir |
| author_facet | Hannan, Abdul Manzoor, Muhammad Arslan Nawaz, Shah Liaqat, Muhammad Irzam Schedl, Markus Noman, Mubashir |
| contents | We study the task of learning association between faces and voices, which is gaining interest in the multimodal community lately. These methods suffer from the deliberate crafting of negative mining procedures as well as the reliance on the distant margin parameter. These issues are addressed by learning a joint embedding space in which orthogonality constraints are applied to the fused embeddings of faces and voices. However, embedding spaces of faces and voices possess different characteristics and require spaces to be aligned before fusing them. To this end, we propose a method that accurately aligns the embedding spaces and fuses them with an enhanced gated fusion thereby improving the performance of face-voice association. Extensive experiments on the VoxCeleb dataset reveals the merits of the proposed approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_17002 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PAEFF: Precise Alignment and Enhanced Gated Feature Fusion for Face-Voice Association Hannan, Abdul Manzoor, Muhammad Arslan Nawaz, Shah Liaqat, Muhammad Irzam Schedl, Markus Noman, Mubashir Computer Vision and Pattern Recognition Artificial Intelligence We study the task of learning association between faces and voices, which is gaining interest in the multimodal community lately. These methods suffer from the deliberate crafting of negative mining procedures as well as the reliance on the distant margin parameter. These issues are addressed by learning a joint embedding space in which orthogonality constraints are applied to the fused embeddings of faces and voices. However, embedding spaces of faces and voices possess different characteristics and require spaces to be aligned before fusing them. To this end, we propose a method that accurately aligns the embedding spaces and fuses them with an enhanced gated fusion thereby improving the performance of face-voice association. Extensive experiments on the VoxCeleb dataset reveals the merits of the proposed approach. |
| title | PAEFF: Precise Alignment and Enhanced Gated Feature Fusion for Face-Voice Association |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2505.17002 |