Multi-Modal Scene Graph with Kolmogorov-Arnold Experts for Audio-Visual Question Answering

TL;DR

This paper introduces a novel multi-modal scene graph and Kolmogorov-Arnold Expert Network to improve audio-visual question answering by better modeling complex interactions and relationships in video content.

Contribution

It presents the first multi-modal scene graph for audio-visual scenes and integrates a Kolmogorov-Arnold Network-based Mixture of Experts for enhanced temporal reasoning.

Findings

Achieves state-of-the-art results on MUSIC-AVQA benchmarks.

Effectively models fine-grained cross-modal interactions.

Improves temporal reasoning in audio-visual question answering.

Abstract

In this paper, we propose a novel Multi-Modal Scene Graph with Kolmogorov-Arnold Expert Network for Audio-Visual Question Answering (SHRIKE). The task aims to mimic human reasoning by extracting and fusing information from audio-visual scenes, with the main challenge being the identification of question-relevant cues from the complex audio-visual content. Existing methods fail to capture the structural information within video, and suffer from insufficient fine-grained modeling of multi-modal features. To address these issues, we are the first to introduce a new multi-modal scene graph that explicitly models the objects and their relationship as a visually grounded, structured representation of the audio-visual scene. Furthermore, we design a Kolmogorov-Arnold Network~(KAN)-based Mixture of Experts (MoE) to enhance the expressive power of the temporal integration stage. This enables more fine-grained modeling of cross-modal interactions within the question-aware fused audio-visual representation, leading to capture richer and more nuanced patterns and then improve temporal reasoning performance. We evaluate the model on the established MUSIC-AVQA and MUSIC-AVQA v2 benchmarks, where it achieves state-of-the-art performance. Code and model checkpoints will be publicly released.