Metraipolis › archive › dc_bfaa9e4fca76

AVC-Reward: Event-Centric Audio-Video Consistency Evaluation — arXiv:2610.22321

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.22321
License (as archived): CC0-1.0
Archived: 2026-10-16T09:31:09.718Z
SHA-256 of the archived text: 087c32a4b86c6fb4643e505a2efa47b113db5ec3a05e4aa52f28f20610b51627
Stable URL of this page: https://metraipolis.com/documento/dc_bfaa9e4fca76

How to cite

AVC-Reward: Event-Centric Audio-Video Consistency Evaluation — arXiv:2610.22321 — arXiv (en), https://arxiv.org/abs/2610.22321 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bfaa9e4fca76 (Metraipolis archive dc_bfaa9e4fca76, sha256 087c32a4b86c6fb4643e505a2efa47b113db5ec3a05e4aa52f28f20610b51627)

BibTeX

@misc{dcbfaa9e4fca76,
  title = {AVC-Reward: Event-Centric Audio-Video Consistency Evaluation — arXiv:2610.22321},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.22321},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bfaa9e4fca76; sha256 087c32a4b86c6fb4643e505a2efa47b113db5ec3a05e4aa52f28f20610b51627},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22321. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As joint audio-video generation models become more capable, they can generate audio-visual scenes exhibiting three forms of complexity: multi-event composition, multi-layer audio mixture, and asymmetric cross-modal relations. However, current evaluations largely reuse metrics from earlier simple and unidirectional tasks. They typically evaluate an audio-video clip as a whole and rely on simple predefined relations, making them unsuitable for these increasingly complex scenes. We propose an event-centric evaluation framework that disentangles audio and visual events at the event level, analyzes each event independently, cross-validates relations across modalities, and aggregates the judgments into a coarse-to-fine, three-level consistency metric. The metric assesses global narrative harmony, event-level temporal synchronization, and fine-grained attribute alignment, providing a general formulation for audio-video consistency evaluation across diverse generation settings and cross-modal relations. We instantiate the metric as AVC-Reward, a trainable audio-video reward model. To evaluate the agreement between automatic judgments and human perception across generation paradigms and relation complexities, we construct AVC-Benchmark, which is primarily built from diverse positive and negative samples generated by various types of audio-video generation models. AVC-Reward achieves state-of-the-art alignment with human judgment among automatic methods, provides interpretable rationales that localize failures, and improves the audio-video consistency of a joint generation model when used as an online reinforcement-learning signal. Project page: this https URL

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.22321 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T08:44:18.055Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).