# Broadband DOA estimation using Convolutional neural networks trained with noise signals

## TL;DR

This paper introduces a CNN-based broadband DOA estimation method that uses phase information from microphone signals, trained with noise signals, and demonstrates robustness and generalization to speech sources and various acoustic conditions.

## Contribution

The paper presents a novel CNN training approach using synthesized noise signals for broadband DOA estimation, simplifying data preparation and enhancing robustness.

## Key findings

- CNN trained with noise signals generalizes well to speech sources
- System is robust to noise and microphone perturbations
- Effective across different acoustic environments

## Abstract

A convolution neural network (CNN) based classification method for broadband DOA estimation is proposed, where the phase component of the short-time Fourier transform coefficients of the received microphone signals are directly fed into the CNN and the features required for DOA estimation are learnt during training. Since only the phase component of the input is used, the CNN can be trained with synthesized noise signals, thereby making the preparation of the training data set easier compared to using speech signals. Through experimental evaluation, the ability of the proposed noise trained CNN framework to generalize to speech sources is demonstrated. In addition, the robustness of the system to noise, small perturbations in microphone positions, as well as its ability to adapt to different acoustic conditions is investigated using experiments with simulated and real data.