Audio-Visual Intelligence in Large Foundation Models

Hugging Face Daily Papers Papers

Summary

This survey paper provides a comprehensive review of audio-visual intelligence within large foundation models, establishing a unified taxonomy, synthesizing core methodologies, and outlining key datasets, benchmarks, and open research challenges.

Audio-Visual Intelligence (AVI) has emerged as a central frontier in artificial intelligence, bridging auditory and visual modalities to enable machines that can perceive, generate, and interact in the multimodal real world. In the era of large foundation models, joint modeling of audio and vision has become increasingly crucial, i.e., not only for understanding but also for controllable generation and reasoning across dynamic, temporally grounded signals. Recent advances, such as Meta MovieGen and Google Veo-3, highlight the growing industrial and academic focus on unified audio-vision architectures that learn from massive multimodal data. However, despite rapid progress, the literature remains fragmented, spanning diverse tasks, inconsistent taxonomies, and heterogeneous evaluation practices that impede systematic comparison and knowledge integration. This survey provides the first comprehensive review of AVI through the lens of large foundation models. We establish a unified taxonomy covering the broad landscape of AVI tasks, ranging from understanding (e.g., speech recognition, sound localization) to generation (e.g., audio-driven video synthesis, video-to-audio) and interaction (e.g., dialogue, embodied, or agentic interfaces). We synthesize methodological foundations, including modality tokenization, cross-modal fusion, autoregressive and diffusion-based generation, large-scale pretraining, instruction alignment, and preference optimization. Furthermore, we curate representative datasets, benchmarks, and evaluation metrics, offering a structured comparison across task families and identifying open challenges in synchronization, spatial reasoning, controllability, and safety. By consolidating this rapidly expanding field into a coherent framework, this survey aims to serve as a foundational reference for future research on large-scale AVI.
Original Article
View Cached Full Text

Cached at: 05/08/26, 02:27 PM

Paper page - Audio-Visual Intelligence in Large Foundation Models

Source: https://huggingface.co/papers/2605.04045 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Audio-Visual Intelligence represents a multidisciplinary field integrating auditory and visual modalities through large foundation models, encompassing tasks from understanding and generation to interaction, with unified taxonomies and methodological foundations.

Audio-Visual Intelligence(AVI) has emerged as a central frontier in artificial intelligence, bridging auditory and visual modalities to enable machines that can perceive, generate, and interact in the multimodal real world. In the era oflarge foundation models, joint modeling of audio and vision has become increasingly crucial, i.e., not only for understanding but also for controllable generation and reasoning across dynamic, temporally grounded signals. Recent advances, such as Meta MovieGen and Google Veo-3, highlight the growing industrial and academic focus on unified audio-vision architectures that learn from massivemultimodal data. However, despite rapid progress, the literature remains fragmented, spanning diverse tasks, inconsistent taxonomies, and heterogeneous evaluation practices that impede systematic comparison and knowledge integration. This survey provides the first comprehensive review of AVI through the lens oflarge foundation models. We establish a unified taxonomy covering the broad landscape of AVI tasks, ranging from understanding (e.g.,speech recognition,sound localization) to generation (e.g.,audio-driven video synthesis,video-to-audio) and interaction (e.g.,dialogue, embodied, oragentic interfaces). We synthesize methodological foundations, includingmodality tokenization,cross-modal fusion, autoregressive anddiffusion-based generation,large-scale pretraining,instruction alignment, andpreference optimization. Furthermore, we curate representative datasets, benchmarks, and evaluation metrics, offering a structured comparison across task families and identifying open challenges insynchronization,spatial reasoning,controllability, andsafety. By consolidating this rapidly expanding field into a coherent framework, this survey aims to serve as a foundational reference for future research on large-scale AVI.

View arXiv pageView PDFProject pageGitHub15Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.04045 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.04045 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.04045 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

LTX-2: Efficient Joint Audio-Visual Foundation Model

Papers with Code Trending

LTX-2 is introduced as an efficient joint audio-visual foundation model. The text includes a mix of the paper reference and a video script about countries facing existential threats, but the primary classification target is the AI model paper.

When Vision Speaks for Sound

Hugging Face Daily Papers

This paper identifies that video-capable multimodal LLMs often appear to understand audio but actually rely on visual cues, a failure mode termed the audio-visual Clever Hans effect. It introduces Thud, an intervention-driven probing framework to diagnose this issue, and proposes an alignment recipe that improves audio-visual consistency by 28 percentage points.

Scalable Visual Pretraining for Language Intelligence

Hugging Face Daily Papers

This paper demonstrates that unsupervised visual pretraining on documents, without text extraction, consistently outperforms text-only pretraining for language intelligence, providing an efficient and scalable approach for foundation models.