DOA Estimation with Lightweight Network on LLM-Aided Simulated Acoustic Scenes

TL;DR

This paper introduces LightDOA, a lightweight neural network for DOA estimation that leverages a novel LLM-assisted dataset to improve robustness and efficiency in diverse acoustic environments.

Contribution

The paper presents LightDOA, a novel low-complexity neural network architecture for DOA estimation, and demonstrates its effectiveness on a new LLM-generated, diverse acoustic scene dataset.

Findings

LightDOA achieves high accuracy across various acoustic scenes.

The model maintains low computational complexity suitable for resource-limited devices.

LLM-assisted dataset enhances the generalizability of DOA models.

Abstract

Direction-of-Arrival (DOA) estimation is critical in spatial audio and acoustic signal processing, with wide-ranging applications in real-world. Most existing DOA models are trained on synthetic data by convolving clean speech with room impulse responses (RIRs), which limits their generalizability due to constrained acoustic diversity. In this paper, we revisit DOA estimation using a recently introduced dataset constructed with the assistance of large language models (LLMs), which provides more realistic and diverse spatial audio scenes. We benchmark several representative neural-based DOA methods on this dataset and propose LightDOA, a lightweight DOA estimation model based on depthwise separable convolutions, specifically designed for mutil-channel input in varying environments. Experimental results show that LightDOA achieves satisfactory accuracy and robustness across various acoustic scenes while maintaining low computational complexity. This study not only highlights the potential of spatial audio synthesized with the assistance of LLMs in advancing robust and efficient DOA estimation research, but also highlights LightDOA as efficient solution for resource-constrained applications.