[Python] pyAudioAnalysis speaker_diarization doesn't work well

from unittest import result import numpy as np import librosa from pydub import AudioSegment from pyAudioAnalysis import audioSegmentation as aS

audio_file_path = "G:\Project\Nunu\duet\test.wav"

signal, sample_rate = librosa.load(audio_file_path, sr=None)

print(f"Sample Rate: {sample_rate}, Signal Shape: {signal.shape}")

segments, labels, unused = aS.speaker_diarization(audio_file_path, n_speakers=2) # n_speakers=0 自动识别

print(f"Segments: {segments}") print(f"Labels: {labels}")

run results in VS code.

Sample Rate: 44100, Signal Shape: (1299867,) Segments: [1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1] Labels: -1

I can play test.wav with ffplay and other media players. why doesn't speaker_diarization work well in my code.

I want to know why speaker_diarization doesn't work well in my code, and how to solve this problem.

Continue reading...