Pushing the Frontier of Audiovisual Perception with Large-Scale Multimodal Correspondence Learning

TL;DR

This paper presents PE-AV, a large-scale multimodal encoder trained with contrastive learning to improve audiovisual perception, enabling new cross-modal tasks and achieving state-of-the-art results across benchmarks.

Contribution

Introduction of PE-AV, a unified audiovisual encoder supporting cross-modal embeddings and large-scale training with synthesized captions for diverse audio-visual data.

Findings

Achieved state-of-the-art performance on standard audio and video benchmarks.

Enabled novel tasks like speech retrieval through unified embeddings.

Improved zero-shot performance by scaling contrastive objectives.

Abstract

We introduce Perception Encoder Audiovisual, PE-AV, a new family of encoders for audio and video understanding trained with scaled contrastive learning. Built on PE, PE-AV makes several key contributions to extend representations to audio, and natively support joint embeddings across audio-video, audio-text, and video-text modalities. PE-AV's unified cross-modal embeddings enable novel tasks such as speech retrieval, and set a new state of the art across standard audio and video benchmarks. We unlock this by building a strong audiovisual data engine that synthesizes high-quality captions for O(100M) audio-video pairs, enabling large-scale supervision consistent across modalities. Our audio data includes speech, music, and general sound effects-avoiding single-domain limitations common in prior work. We exploit ten pairwise contrastive objectives, showing that scaling cross-modality and caption-type pairs strengthens alignment and improves zero-shot performance. We further develop PE-A-Frame by fine-tuning PE-AV with frame-level contrastive objectives, enabling fine-grained audio-frame-to-text alignment for tasks such as sound event detection.