Towards Universal Video MLLMs with Attribute-Structured and Quality-Verified Instructions

TL;DR

This paper introduces a large, structured audiovisual instruction dataset and a new video understanding model that enhances fine-grained captioning and instruction following, advancing universal video understanding.

Contribution

The paper presents ASID-1M, a comprehensive dataset with attribute-structured annotations, a verification pipeline, and a new model trained on this data to improve audiovisual captioning.

Findings

Achieves state-of-the-art results among open-source models

Reduces hallucinations in captions

Improves fine-grained caption quality

Abstract

Universal video understanding requires modeling fine-grained visual and audio information over time in diverse real-world scenarios. However, the performance of existing models is primarily constrained by video-instruction data that represents complex audiovisual content as single, incomplete descriptions, lacking fine-grained organization and reliable annotation. To address this, we introduce: (i) ASID-1M, an open-source collection of one million structured, fine-grained audiovisual instruction annotations with single- and multi-attribute supervision; (ii) ASID-Verify, a scalable data curation pipeline for annotation, with automatic verification and refinement that enforces semantic and temporal consistency between descriptions and the corresponding audiovisual content; and (iii) ASID-Captioner, a video understanding model trained via Supervised Fine-Tuning (SFT) on the ASID-1M. Experiments across seven benchmarks covering audiovisual captioning, attribute-wise captioning, caption-based QA, and caption-based temporal grounding show that ASID-Captioner improves fine-grained caption quality while reducing hallucinations and improving instruction following. It achieves state-of-the-art performance among open-source models and is competitive with Gemini-3-Pro.