audio-kirtan-recognizer

Audio Kirtan Recognizer

Note: This repo started during a weekend hackathon, so the quality needs to be improved

authenticate w/ google cloud

gcloud auth application-default login

Note: APIs for Automatic speech to text should be enabled

upload audio files to google cloud storage

• get gs_urls of the audio files
• ghe format will be something like gs://my_bucket_name//path_to_audio_file
• edit the main.py file and run it

python main.py

converting mp3 to flac

ffmpeg -i input.mp3 output.flac

Hackathon Collab links

Audio File Segmentation

Music Source separation

ASR Transcription to BaniDB