Delayed Commitment for Representation Readiness in Stage-wise Audio-Visual Learning

TL;DR

This paper introduces DPC-Net, a framework that improves stage-wise audio-visual learning by estimating and correcting representation readiness, leading to better performance across multiple tasks.

Contribution

It formulates the readiness deficiency problem and proposes DPC-Net to localize and correct bottlenecks in representation propagation for audio-visual tasks.

Findings

DPC-Net improves performance in speech separation, event localization, and speech recognition.

The method effectively localizes intervention-sensitive bottlenecks.

Readiness-guided correction enhances the quality of fused representations.

Abstract

Stage-wise audio-visual encoders propagate fused intermediate states across layers, making the formation of later representations depend on the readiness of earlier fusion states. Strong local audio-visual agreement provides useful correspondence evidence, yet a fused state also needs sufficient cross-layer and cross-modal support before it can reliably guide later fusion. This paper studies this issue through propagation-aware representation readiness and formulates premature perceptual commitment as a readiness-deficiency problem, where local plausibility, propagation influence, and support insufficiency jointly appear at an intermediate stage. We propose the Delayed Perceptual Commitment Network (DPC-Net), an encoder-level framework that estimates an observable readiness-deficiency surrogate, localizes the intervention-sensitive bottleneck, and applies support-aware correction with cross-layer and cross-modal evidence. DPC-Net preserves task-specific heads, losses, decoding modules, and evaluation protocols, making it applicable to different audio-visual tasks through encoder-side intervention. Experiments on audio-visual speech separation, audio-visual event localization, and audio-visual speech recognition show consistent improvements across reconstruction, localization, and recognition regimes. Further analyses on component contribution, selection criteria, counterfactual intervention, and readiness trajectories support the effectiveness of readiness-guided bottleneck correction.