SPUR: A Plug-and-Play Framework for Integrating Spatial Audio Understanding and Reasoning into Large Audio-Language Models

2025/11/10 by S Sakshi, Vaibhavi Lokegaonkar, Sakshi, S +10 · 1 citation
Computer Science · Psychology · #Artificial Intelligence (cs.AI) #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Music and Audio Processing #Phonetics and Phonology Research #Speech and Audio Processing #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2511.06606

openalex publication_date 2025/11/10 · openalex created_date 2025/11/12 · openalex updated_date 2026/07/28

Abstract

Spatial perception is central to auditory intelligence, enabling accurate understanding of real-world acoustic scenes and advancing human-level perception of the world around us. While recent large audio-language models (LALMs) show strong reasoning over complex audios, most operate on monaural inputs and lack the ability to capture spatial cues such as direction, elevation, and distance. We introduce SPUR, a lightweight, plug-in approach that equips LALMs with spatial perception through minimal architectural changes. SPUR consists of: (i) a First-Order Ambisonics (FOA) encoder that maps (W, X, Y, Z) channels to rotation-aware, listener-centric spatial features, integrated into target LALMs via a multimodal adapter; and (ii) SPUR-Set, a spatial QA dataset combining open-source FOA recordings with controlled simulations, emphasizing relative direction, elevation, distance, and overlap for supervised spatial reasoning. Fine-tuning our model on the SPUR-Set consistently improves spatial QA and multi-speaker attribution while preserving general audio understanding. SPUR provides a simple recipe that transforms monaural LALMs into spatially aware models. Extensive ablations validate the effectiveness of our approach.

Citations

• OWL: Geometry-Aware Spatial Reasoning for Audio Large Language Models
• SALM: Spatial Audio Language Model with Structured Embeddings for Understanding and Editing
• Step-Audio 2 Technical Report
• Towards Spatial Audio Understanding via Question Answering
• Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
• Spatial Audio Processing with Large Language Model on Wearable Devices
• Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities
• Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs
• Blind Estimation of Sub-band Acoustic Parameters from Ambisonics Recordings using Spectro-Spatial Covariance Features
• GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities
• Can Large Language Models Understand Spatial Audio?
• Audio Flamingo: A Novel Audio Language Model with Few-Shot Learning and Dialogue Abilities
• STARSS23: An Audio-Visual Dataset of Spatial Recordings of Real Scenes with Spatiotemporal Annotations of Sound Events
• Pengi: An Audio Language Model for Audio Tasks
• GPT-4 Technical Report
• Robust Speech Recognition via Large-Scale Weak Supervision
• STARSS22: A dataset of spatial recordings of real scenes with spatiotemporal annotations of sound events
• LoRA: Low-Rank Adaptation of Large Language Models
• A Dataset of Dynamic Reverberant Sound Scenes with Directional Interferers for Sound Event Localization and Detection

Cited by

• A Survey of Large Audio Language Models: Generalization, Trustworthiness, and Outlook

Related

• BAT: Learning to Reason about Spatial Sounds with Large Language Models
• OWL: Geometry-Aware Spatial Reasoning for Audio Large Language Models
• Can Large Language Models Understand Spatial Audio?
• SAVVY: Spatial Awareness via Audio-Visual LLMs through Seeing and Hearing
• Spatial Blind Spot: Auditory Motion Perception Deficits in Audio LLMs