/compare20-MB

ComParE 2020 Mask and Breathing Subchallenges

Primary LanguageJupyter Notebook

Ensembling End-to-End Deep Models for Computational Paralinguistics Tasks: ComParE 2020 Mask and Breathing Sub-Challenges

ComParE 2020 Mask and Breathing Sub-Challenges

Description

This project contains the scripts for our entry in the Mask and Breathing Sub-Challenges, which are the part of the INTERSPEECH 2020 Computational Paralinguistics Challenge (ComParE 2020).

The pipeline of proposed system in the Mask Sub-Challenge

1 \

The pipeline of proposed system in the Breathing Sub-Challenge

2

Project Organization

.
├── LICENSE
├── README.md                           <- README for developers using this project.
│
├── Breathing                           <- Python project dedicated to breathing recognizing with the help of deep neural network (end-to-end) models. 
|                                          Please see Breathing/README.md for more details.
|
├── Mask                                <- Python project dedicated to recognizing the mask on the face from recorded audio by different pretrained deep neural networks.
                                           Please see Breathing/README.md for more details.

For more information, please, contact:
m.markitantov@yandex.ru - Mask Sub-Challenge
denis.dresvyanskiy@uni-ulm.de - Breathing Sub-challenge

Citing

Please cite our paper if you use any code of our solution for ComParE2020 in your research work.
M. Markitantov, D. Dresvyanskiy, D. Mamontov, H. Kaya, W. Minker, and A. Karpov, “Ensembling end-to-end deep modelsfor computational paralinguistics tasks: ComParE 2020 Mask andBreathing Sub-challenges,” in INTERSPEECH, Shanghai, China, October 2020, to appear.