Distribution-Alignment Bridge for Text-Video Retrieval = 텍스트-비디오 검색을 위한 분포 정렬 브릿지

다국어 초록 (Multilingual Abstract)

This paper proposes the Distribution-Alignment Bridge (DAB), a framework that reconceptualizes text-to-video retrieval as a distribution alignment task rather than traditional deterministic point matching. By modeling both text and video embeddings as Gaussian distributions defined by mean and variance, DAB explicitly accounts for modality-specific uncertainty. We employ a deterministic, diffusion-inspired bridge to iteratively refine text distributions toward their target video distributions via a truncated process. This approach unifies probabilistic embedding and distributional transformation into a cohesive, end-to-end trainable system. To optimize cross-modal similarity, we introduce a distribution-aware contrastive loss utilizing Kullback-Leibler divergence. Extensive evaluations on MSR-VTT, MSVD, and VATEX benchmarks confirm that DAB significantly outperforms existing probabilistic and diffusion-based baselines, offering superior retrieval accuracy alongside interpretable uncertainty modeling.

목차 (Table of Contents)

• 1. Introduction 1
• 2. Related Work 4
• 3. Methods 7
• 4. Experiments 19
• 5. Conclusions 28
• References 29
• 텍스트-비디오 검색을 위한 확률적 확산 브릿지 33