This repository contains the code for the paper "Deep complex-valued convolutional-recurrent networks for single source doa estimation" to be published at the International Workshop on Acoustic Signal Enhancement (IWAENC) 2022.
https://hal.science/hal-03779970/document https://ieeexplore.ieee.org/abstract/document/9914747
To test the code without installing anything, we suggest running it using this Kaggle notebook. To install it locally, follow the instructions below.
- Python 3
run pip install -r requirements.txt to install the python libraries needed
Download the Kaggle dataset containing the data, and change the file 'config/dcase_2019_task3_dataset.yaml' to point at the correct train, validation and test datasets.
Then, change the working directory to this project and run python train.py or make train to start training the model. Every time you start training a model, a folder will be created in the outputs/
To execute all unit tests, run either:
pytest tests
or
make tests
`