OmniJigsaw: Enhancing Omni-Modal Reasoning via Modality-Orchestrated Reordering

TL;DR

OmniJigsaw is a self-supervised framework that improves omni-modal reasoning by using a novel temporal reordering proxy task with strategic modality orchestration and data filtering.

Contribution

It introduces a generic self-supervised approach with three modality strategies and a two-stage filtering pipeline for effective omni-modal model training.

Findings

Significant improvements on 15 benchmarks across video, audio, and reasoning tasks.

Mitigates the bi-modal shortcut phenomenon with clip-level modality masking.

Outperforms existing methods in cross-modal understanding and reasoning.

Abstract

To extend the reinforcement learning post-training paradigm to omni-modal models for concurrently bolstering video-audio understanding and collaborative reasoning, we propose OmniJigsaw, a generic self-supervised framework built upon a temporal reordering proxy task. Centered on the chronological reconstruction of shuffled audio-visual clips, this paradigm strategically orchestrates visual and auditory signals to compel cross-modal integration through three distinct strategies: Joint Modality Integration, Sample-level Modality Selection, and Clip-level Modality Masking. Recognizing that the efficacy of such proxy tasks is fundamentally tied to puzzle quality, we design a two-stage coarse-to-fine data filtering pipeline, which facilitates the efficient adaptation of OmniJigsaw to massive unannotated omni-modal data. Our analysis reveals a ``bi-modal shortcut phenomenon'' in joint modality integration and demonstrates that fine-grained clip-level modality masking mitigates this issue while outperforming sample-level modality selection. Extensive evaluations on 15 benchmarks show substantial gains in video, audio, and collaborative reasoning, validating OmniJigsaw as a scalable paradigm for self-supervised omni-modal learning.