Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2512.24731 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917178019676160 |
|---|---|
| author | Li, Bingxuan Cui, Yiming He, Yicheng Wang, Yiwei Zhang, Shu Wen, Longyin Niu, Yulei |
| author_facet | Li, Bingxuan Cui, Yiming He, Yicheng Wang, Yiwei Zhang, Shu Wen, Longyin Niu, Yulei |
| contents | Sound effects build an essential layer of multimodal storytelling, shaping the emotional atmosphere and the narrative semantics of videos. Despite recent advancement in video-text-to-audio (VT2A), the current formulation faces three key limitations: First, an imbalance between visual and textual conditioning that leads to visual dominance; Second, the absence of a concrete definition for fine-grained controllable generation; Third, weak instruction understanding and following, as existing datasets rely on brief categorical tags. To address these limitations, we introduce EchoFoley, a new task designed for video-grounded sound generation with both event level local control and hierarchical semantic control. Our symbolic representation for sounding events specifies when, what, and how each sound is produced within a video or instruction, enabling fine-grained controls like sound generation, insertion, and editing. To support this task, we construct EchoFoley-6k, a large-scale, expert-curated benchmark containing over 6,000 video-instruction-annotation triplets. Building upon this foundation, we propose EchoVidia a sounding-event-centric agentic generation framework with slow-fast thinking strategy. Experiments show that EchoVidia surpasses recent VT2A models by 40.7% in controllability and 12.5% in perceptual quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_24731 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | EchoFoley: Event-Centric Hierarchical Control for Video Grounded Creative Sound Generation Li, Bingxuan Cui, Yiming He, Yicheng Wang, Yiwei Zhang, Shu Wen, Longyin Niu, Yulei Computer Vision and Pattern Recognition Sound effects build an essential layer of multimodal storytelling, shaping the emotional atmosphere and the narrative semantics of videos. Despite recent advancement in video-text-to-audio (VT2A), the current formulation faces three key limitations: First, an imbalance between visual and textual conditioning that leads to visual dominance; Second, the absence of a concrete definition for fine-grained controllable generation; Third, weak instruction understanding and following, as existing datasets rely on brief categorical tags. To address these limitations, we introduce EchoFoley, a new task designed for video-grounded sound generation with both event level local control and hierarchical semantic control. Our symbolic representation for sounding events specifies when, what, and how each sound is produced within a video or instruction, enabling fine-grained controls like sound generation, insertion, and editing. To support this task, we construct EchoFoley-6k, a large-scale, expert-curated benchmark containing over 6,000 video-instruction-annotation triplets. Building upon this foundation, we propose EchoVidia a sounding-event-centric agentic generation framework with slow-fast thinking strategy. Experiments show that EchoVidia surpasses recent VT2A models by 40.7% in controllability and 12.5% in perceptual quality. |
| title | EchoFoley: Event-Centric Hierarchical Control for Video Grounded Creative Sound Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.24731 |