Audio Analyzer is a pipeline for audio analysis that aims to provide a comprehensive solution for audio processing. The project is designed to be modular and scalable, allowing users to easily add new features and functionalities.
- create a directory for each run to debug with time stamp
- check on large audio files
- split to 1 minute segments
- extract features from each segment [make it parallel]
- save features to csv
- not assuming language
- background noise classification
- audio enhancement
- speaker diarization Link1, Link2
- tone classification
- text extraction
- sentiment analysis
- toxic words detection
- text summarization
- cleaning dead segments
- audio segmentation - part of day
- run pipreqs to generate requirements.txt
pipreqs . --force
- run:
pip install -r requirements.txt
bugfixes:
-
https://stackoverflow.com/questions/67069960/pydub-error-loading-file-unknown-encoder-pcm-s4le
code:
import soundfile as sf #Load the original audio file (in IMA ADPCM format) original_audio, sample_rate = sf.read('original.wav') #Save the audio with the desired format (PCM_16) sf.write('output.wav', original_audio, sample_rate, subtype='PCM_16')
-
code:
bash ~/Downloads/Miniconda3-latest-MacOSX-arm64.sh -b -p $HOME/miniconda source ~/miniconda/bin/activate conda install -c apple tensorflow-deps