Learning Audio-Visual Embeddings with Inferred Latent Interaction Graphs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zeng, Donghuo, Niu, Hao, Wang, Yanan, Taya, Masato
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911382588358656
author Zeng, Donghuo
Niu, Hao
Wang, Yanan
Taya, Masato
author_facet Zeng, Donghuo
Niu, Hao
Wang, Yanan
Taya, Masato
contents Learning robust audio-visual embeddings requires bringing genuinely related audio and visual signals together while filtering out incidental co-occurrences - background noise, unrelated elements, or unannotated events. Most contrastive and triplet-loss methods use sparse annotated labels per clip and treat any co-occurrence as semantic similarity. For example, a video labeled "train" might also contain motorcycle audio and visual, because "motorcycle" is not the chosen annotation; standard methods treat these co-occurrences as negatives to true motorcycle anchors elsewhere, creating false negatives and missing true cross-modal dependencies. We propose a framework that leverages soft-label predictions and inferred latent interactions to address these issues: (1) Audio-Visual Semantic Alignment Loss (AV-SAL) trains a teacher network to produce aligned soft-label distributions across modalities, assigning nonzero probability to co-occurring but unannotated events and enriching the supervision signal. (2) Inferred Latent Interaction Graph (ILI) applies the GRaSP algorithm to teacher soft labels to infer a sparse, directed dependency graph among classes. This graph highlights directional dependencies (e.g., "Train (visual)" -> "Motorcycle (audio)") that expose likely semantic or conditional relationships between classes; these are interpreted as estimated dependency patterns. (3) Latent Interaction Regularizer (LIR): A student network is trained with both metric loss and a regularizer guided by the ILI graph, pulling together embeddings of dependency-linked but unlabeled pairs in proportion to their soft-label probabilities. Experiments on AVE and VEGAS benchmarks show consistent improvements in mean average precision (mAP), demonstrating that integrating inferred latent interactions into embedding learning enhances robustness and semantic coherence.
format Preprint
id arxiv_https___arxiv_org_abs_2601_11995
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning Audio-Visual Embeddings with Inferred Latent Interaction Graphs
Zeng, Donghuo
Niu, Hao
Wang, Yanan
Taya, Masato
Multimedia
Artificial Intelligence
Information Retrieval
Machine Learning
Sound
Learning robust audio-visual embeddings requires bringing genuinely related audio and visual signals together while filtering out incidental co-occurrences - background noise, unrelated elements, or unannotated events. Most contrastive and triplet-loss methods use sparse annotated labels per clip and treat any co-occurrence as semantic similarity. For example, a video labeled "train" might also contain motorcycle audio and visual, because "motorcycle" is not the chosen annotation; standard methods treat these co-occurrences as negatives to true motorcycle anchors elsewhere, creating false negatives and missing true cross-modal dependencies. We propose a framework that leverages soft-label predictions and inferred latent interactions to address these issues: (1) Audio-Visual Semantic Alignment Loss (AV-SAL) trains a teacher network to produce aligned soft-label distributions across modalities, assigning nonzero probability to co-occurring but unannotated events and enriching the supervision signal. (2) Inferred Latent Interaction Graph (ILI) applies the GRaSP algorithm to teacher soft labels to infer a sparse, directed dependency graph among classes. This graph highlights directional dependencies (e.g., "Train (visual)" -> "Motorcycle (audio)") that expose likely semantic or conditional relationships between classes; these are interpreted as estimated dependency patterns. (3) Latent Interaction Regularizer (LIR): A student network is trained with both metric loss and a regularizer guided by the ILI graph, pulling together embeddings of dependency-linked but unlabeled pairs in proportion to their soft-label probabilities. Experiments on AVE and VEGAS benchmarks show consistent improvements in mean average precision (mAP), demonstrating that integrating inferred latent interactions into embedding learning enhances robustness and semantic coherence.
title Learning Audio-Visual Embeddings with Inferred Latent Interaction Graphs
topic Multimedia
Artificial Intelligence
Information Retrieval
Machine Learning
Sound
url https://arxiv.org/abs/2601.11995