评测四大音视频同步度量，发现无绝对优劣，需分场景使用。
What Do Audio-Visual Synchronization Metrics Actually Measure?
- 构建统一评测框架，检验各度量在扰动下的稳定性与一致性。
- Synchformer/DeSync最擅长捕捉时间偏移（τ=0.84），而ImageBind/JavisScore更贴近人类判断。
- 各度量相互矛盾（Krippendorff α=0.066），融合也难提升表现，建议报告可靠性卡片。
自动音视频同步度量广泛用于排序和训练生成模型，但很少作为测量工具被系统评估。本文采用统一可靠性协议，联合审计 AV-Align、ImageBind AV-relevance、JavisScore 以及 Synchformer/DeSync：包括受控畸变单调性、预处理敏感性、排名不确定性、跨度量一致性、与 PEAVS 人类对齐代理的一致性，以及学习融合能力。结果呈现分裂轴线而非单一胜者：Synchformer/DeSync 在时间偏移追踪上最强（τ=0.84），ImageBind/JavisScore 更匹配人类判断与内容中断情形（τ=0.20），而 AV-Align 是最弱的独立度量。各度量间存在显著分歧（Krippendorff α=0.066），且线性或简单 k-NN 融合无法超越最优单个度量的 PEAVS 一致性。建议将音视频同步评估以‘可靠性卡片’形式报告，包含度量家族分解与置信区间，而非单一分数。
原文摘要 · Abstract (English)
Automatic AV-sync metrics are widely used to rank and train audio-visual generators, but they are rarely audited as measurement instruments. We jointly audit AV-Align, ImageBind AV-relevance, JavisScore, and Synchformer/DeSync under a common reliability protocol: controlled-distortion monotonicity, preprocessing sensitivity, rank uncertainty, cross-metric agreement, PEAVS-proxy agreement, and learned fusion. The result is an axis split, not a single winner: Synchformer/DeSync is the strongest temporal-offset tracker ($τ=0.84$), ImageBind/JavisScore better match the PEAVS human-aligned proxy ($τ=0.20$) and content-disruption families, and AV-Align is the weakest standalone metric. The metrics mutually disagree (Krippendorff $α=0.066$), and neither linear nor simple $k$-NN fusion improves PEAVS agreement over the best individual metric. We recommend reporting AV-sync as a Reliability Card (metric-family breakdowns with confidence intervals) rather than a single bare synchronization score.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。