This repository contains the codebase accompanying our publication:
Christopher Schymura, Benedikt Bönninghoff, Tsubasa Ochiai, Marc Delcroix, Keisuke Kinoshita, Tomohiro Nakatani, Shoko Araki, Dorothea Kolossa, "PILOT: Introducing Transformers for Probabilistic Sound Event Localization", INTERSPEECH 2021
[ arXiv ]
Sound event localization aims at estimating the positions of sound sources in the environment with respect to an acoustic receiver (e.g. a microphone array). Recent advances in this domain most prominently focused on utilizing deep recurrent neural networks. Inspired by the success of transformer architectures as a suitable alternative to classical recurrent neural networks, the PILOT (Probabilistic Localization of Sounds with Transformers) model is a transformer-based sound event localization framework, where temporal dependencies in the received multi-channel audio signals are captured via self-attention mechanisms. Additionally, the estimated sound event positions are represented as multivariate Gaussian variables, yielding an additional notion of uncertainty, which many previously proposed deep learning-based systems designed for this application do not provide. The general architecture of PILOT is shown in the figure below.
You can train and evaluate the PILOT model using the ANSIM, RESIM and REAL sound event localization and detection datasets. We have prepared a script that downloads the respective datasets and stores them in a suitable folder structure. Simply run
$ ./download_data.sh dataset-name
where dataset-name
specifies the desired dataset (either ansim
, resim
or real
).