SELD-TCN: Sound Event Localization and Detection via Temporal Convolutional Networks

This repository contains a SELD-TCN implementation. The implementation is discussed here.

The SELD-TCN was originally proposed in https://arxiv.org/abs/2003.01609.
SELDnet implementation (https://github.com/sharathadavanne/seld-net.git) has been used as a starting point for the development of this network.

Dataset: ANSIM (TUT Sound Events 2018 - Ambisonic, Anechoic and Synthetic Impulse Response Dataset)

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
MATLAB_preprocessing		MATLAB_preprocessing
__pycache__		__pycache__
images		images
img		img
models		models
LICENSE		LICENSE
README.md		README.md
SELD-TCN_implementation_report.pdf		SELD-TCN_implementation_report.pdf
batch_feature_extraction.py		batch_feature_extraction.py
cls_data_generator.py		cls_data_generator.py
cls_feature_class.py		cls_feature_class.py
evaluation_metrics.py		evaluation_metrics.py
keras_model.py		keras_model.py
parameter.py		parameter.py
seld.py		seld.py
utils.py		utils.py

Provide feedback