{"repo":"Voice-Privacy-Challenge/Voice-Privacy-Challenge-2022","free":true,"listed":false,"github":"https://github.com/Voice-Privacy-Challenge/Voice-Privacy-Challenge-2022","clone":"git clone https://github.com/Voice-Privacy-Challenge/Voice-Privacy-Challenge-2022.git","description":"Baseline Recipe for VoicePrivacy Challenge 2022: anonymization systems and evaluation software","language":"Python","stars":69,"topics":["anonymization","speaker-recognition","asr","privacy-protection","voice-privacy-challenge","voice-privacy","voice-anonymization","privacy","attack-model","anonymization-metrics"],"license":null,"category":"analytics","readme_excerpt":"Recipe for VoicePrivacy Challenge 2022 Please visit the challenge website for more information about the Challenge. Install 1. git clone --recurse-submodules https://github.com/Voice-Privacy-Challenge/Voice-Privacy-Challenge-2022.git 2. ./install.sh Running the recipe The recipe uses the pre-trained models of anonymization. To run the baseline system with evaluation: 1. cd baseline 2. run ./run.sh . In run.sh, to download models and data the user will be requested the password which is provided during the Challenge registration. Re-running the scripts for anonymization / evaluation Use cleanup.sh to remove old data. Check Evaluation for more details. General information For more details about the baseline and data, please see The VoicePrivacy 2022 Challenge Evaluation Plan For the latest updates in the baseline and evaluation scripts, please visit News and updates page The VoicePrivacy 2022 Challenge is over. To get access to evaluation datasets and models, please send an email to organisers@lists.voiceprivacychallenge.org with “VoicePrivacy-2022 registration\" as the subject line. The mail body should include: (i) the contact person; (ii) affiliation; (iii) country; (iv) status (academic/nonacademic). Data Training data The dataset for anonymization system training consists of subsets from the following corpora : LibriSpeech - train-clean-100, train-other-500 LibriTTS - train-clean-100, train-other-500 VoxCeleb 1 & 2 - all only specified subsets of these corpora can be used f","default_branch":null,"files":null,"tree":[],"storefront":"/r/Voice-Privacy-Challenge","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/Voice-Privacy-Challenge/Voice-Privacy-Challenge-2022/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}