System 2 package

Submodules

yt_audio_collector.system_2.preprocess_audio module

Splitting audio based on transcript

class yt_audio_collector.system_2.preprocess_audio.PreProcessAudio(source_path: str = '/home/runner/work/yt_audio_collector/yt_audio_collector/output', destination_path: str = '/home/runner/work/yt_audio_collector/yt_audio_collector/processed_output', background_sound: bool = False)

Bases: object

This class is responsible for preprocessing the audio for speech recognition model. It extracts vocals from audio files, resamples the audio, and divides the audio into chunks based on transcriptions.

extract_vocals(chunk_path: str) → None

Extracts vocals from audio file using spleeter library and store it in the temporary file(temp).

Parameters:

chunk_path: str

The path of the audio file.

static get_file_name(total_file_path: Path) → str

Finds the file name without extension from absolute file path

Parameters:

total_file_path: Path

The absolute path of the file

Return:

str

File name without extension

preprocess_audio()

Traverses through all categories of audio and preprocess the audio.

preprocess_audio_chunks(category_path: str) → None

Divides the audio into chunks based on the transcriptions and preprocess the audio chunks.

Parameters:

category_path: str

The category name in the audio folder.

resample(chunk_path: str, destination_chunk_path: str) → None

Changes the sample rate, sample width, channels of the the given audio file .

Parameters:

chunk_path: str

The audio file path before resampling.

destination_chunk_path: str

The audio fille path after resampling.

Module contents