TF-CADE：面向零样本时序动作检测的前景集中文本-视频对齐
TF-CADE: Foreground-Concentrated Text-Video Alignment for Zero-Shot Temporal Action Detection
提出前景集中的文本-视频对齐方法，通过动作集中聚合和基于确定性的置信度重加权，提升零样本时序动作检测中动作类别的语义区分度，可复用于其他需要细粒度文本-视频对齐的视频理解任务。
原文依据 we propose a Text-Foreground Concentrated Alignment for zero-shot temporal action DEtector (TF-CADE) that explicitly aligns textual information with action-relevant foreground regions.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论零样本时序动作检测（ZSTAD）旨在从未修剪视频中定位并识别未见动作类别的动作实例。尽管现有方法通过改进架构层面的文本-视频对齐已展现出有效性，但它们仍难以捕捉动作类别间的语义区分，导致与文本无关的预测。为解决此问题，我们提出了一种面向零样本时序动作检测器的文本-前景集中对齐方法（TF-CADE），显式地将文本信息与动作相关的前景区域对齐。具体而言，我们引入了动作集中聚合（ACA），提取动作集中分数以将时序信息丰富的视频片段聚合为前景加权视频嵌入。这种前景集中对齐增强了文本与视频特征间的语义一致性，并提升了类间区分度。此外，基于确定性的置信度重加权（CCR）策略利用前景感知相似度细化逐片段置信度分数，在推理时有效抑制无关动作类别。大量评估表明，我们的TF-CADE不仅在分布内设定下达到最优性能，在跨数据集泛化至未见动作类别方面也表现出色。
英文原文摘要 1332 字符
Zero-Shot Temporal Action Detection (ZSTAD) aims to localize and recognize action instances from unseen action categories in untrimmed videos. Although existing methods have shown effectiveness by advancing architectural text-video alignment, they still struggle with capturing semantic distinctions between action classes, resulting in text-irrelevant predictions.To address this issue, we propose a Text-Foreground Concentrated Alignment for zero-shot temporal action DEtector (TF-CADE) that explicitly aligns textual information with action-relevant foreground regions.Specifically, we introduce Action Concentrate Aggregation (ACA), which extracts action concentrate scores to aggregate temporally informative video segments into a foreground-weighted video embedding.This foreground concentrated alignment enhances the semantic consistency between text and video features and improves inter-class discriminability.In addition, a Certainty-based Confidence Re-weighting (CCR) strategy refines per-snippet confidence scores by leveraging foreground-aware similarity, effectively suppressing irrelevant action classes during inference.Extensive evaluations show that our TF-CADE not only achieves state-of-the-art performance under in-distribution settings but also excels in cross-dataset generalization to unseen action classes.