From Sounds to Scenes: A Benchmark for Evaluating Context-Aware Auditory Scene Understanding in Large Audio Language Models

2026-06-24

AI Summary

Current evaluations of large audio language models are largely confined to isolated acoustic layers, failing to assess their holistic understanding of contextual relationships in real-world multi-source auditory scenes. This work proposes the Context-Aware Auditory Scene Understanding (CASU) benchmark, which introduces a scalable, multi-dimensional evaluation framework by integrating speech, acoustic events, and background environments into semi-synthetic audio streams. CASU encompasses four tasks: contextual question answering, entity extraction, speaker role inference, and counterfactual reasoning. Experimental results demonstrate that reliance on a single acoustic modality severely limits model performance, whereas the integration of multi-layered acoustic information is essential for effective auditory scene understanding, thereby validating CASU’s pivotal role in advancing the evaluation of complex audio comprehension capabilities.

Abstract

Recent Large Audio Language Models (LALMs) have achieved remarkable progress in audio perceptual tasks across individual acoustic layers, including speech, sound, and music. However, existing benchmarks predominantly evaluate these layers in isolation, overlooking the complex contextual relationships that arise when multiple acoustic sources co-occur in real-world auditory scenes. Real-world auditory interpretation requires Context-Aware Auditory Scene Understanding (CASU): the ability to comprehend the holistic scene by integrating sound layers. To evaluate this capability, we introduce the CASU benchmark, which assesses whether Audio LLMs can interpret auditory scenes composed of speech, acoustic events (e.g., announcements), and background environments (e.g., traffic), and reason about the logical relationships between these layers. We propose a scalable pipeline for constructing time-accurate, semi-synthetic audio streams by composing real-world scene sounds with synthetic speech. Building on this data, we design four tasks that probe scene understanding: contextual question answering, entity extraction from the scene, speaker role inference, and counterfactual reasoning where scene is manipulated. Experiments across multiple LALMs demonstrate that effective auditory scene understanding requires integration over all auditory layers, rather than reliance on speech or sound alone, underscoring the necessity of CASU for advancing complex audio understanding in LALMs.

Problem

• Context-Aware Auditory Scene Understanding
• Large Audio Language Models
• auditory scenes
• contextual relationships
• audio benchmark

Innovation

• Context-Aware Auditory Scene Understanding
• Large Audio Language Models
• Semi-synthetic Audio Benchmark
• Multi-layer Audio Integration
• Auditory Scene Reasoning