Systematic Evaluation of Time-Frequency Features for Binaural Sound Source Localization

TL;DR

This paper systematically evaluates how different time-frequency features affect binaural sound source localization performance, highlighting the importance of feature selection over model complexity for better generalization.

Contribution

It provides a comprehensive analysis of feature combinations for CNN-based binaural SSL, offering practical guidance on feature design for diverse conditions.

Findings

ILD + IPD features suffice for in-domain SSL

Rich feature sets improve out-of-domain generalization

Low-complexity CNN achieves competitive results with optimal features

Abstract

This study presents a systematic evaluation of time-frequency feature design for binaural sound source localization (SSL), focusing on how feature selection influences model performance across diverse conditions. We investigate the performance of a convolutional neural network (CNN) model using various combinations of amplitude-based features (magnitude spectrogram, interaural level difference - ILD) and phase-based features (phase spectrogram, interaural phase difference - IPD). Evaluations on in-domain and out-of-domain data with mismatched head-related transfer functions (HRTFs) reveal that carefully chosen feature combinations often outperform increases in model complexity. While two-feature sets such as ILD + IPD are sufficient for in-domain SSL, generalization to diverse content requires richer inputs combining channel spectrograms with both ILD and IPD. Using the optimal feature sets, our low-complexity CNN model achieves competitive performance. Our findings underscore the importance of feature design in binaural SSL and provide practical guidance for both domain-specific and general-purpose localization.