metadata
inference: true
pipeline_tag: audio-classification
tags:
- speech
- audio
- HUBert
Working example of using pretrained model to predict emotion in local audio file
def predict_emotion_hubert(audio_file):
""" inspired by an example from https://github.com/m3hrdadfi/soxan """
from audio_models import HubertForSpeechClassification
from transformers import Wav2Vec2FeatureExtractor, AutoConfig
import torch.nn.functional as F
import torch
import numpy as np
from pydub import AudioSegment
model = HubertForSpeechClassification.from_pretrained("Rajaram1996/Hubert_emotion") # Downloading: 362M
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained("facebook/hubert-base-ls960")
sampling_rate=16000 # defined by the model; must convert mp3 to this rate.
config = AutoConfig.from_pretrained("Rajaram1996/Hubert_emotion")
def speech_file_to_array(path, sampling_rate):
# using torchaudio...
# speech_array, _sampling_rate = torchaudio.load(path)
# resampler = torchaudio.transforms.Resample(_sampling_rate, sampling_rate)
# speech = resampler(speech_array).squeeze().numpy()
sound = AudioSegment.from_file(path)
sound = sound.set_frame_rate(sampling_rate)
sound_array = np.array(sound.get_array_of_samples())
return sound_array
sound_array = speech_file_to_array(audio_file, sampling_rate)
inputs = feature_extractor(sound_array, sampling_rate=sampling_rate, return_tensors="pt", padding=True)
inputs = {key: inputs[key].to("cpu").float() for key in inputs}
with torch.no_grad():
logits = model(**inputs).logits
scores = F.softmax(logits, dim=1).detach().cpu().numpy()[0]
outputs = [{
"emo": config.id2label[i],
"score": round(score * 100, 1)}
for i, score in enumerate(scores)
]
return [row for row in sorted(outputs, key=lambda x:x["score"], reverse=True) if row['score'] != '0.0%'][:2]
result = predict_emotion_hubert("male-crying.mp3")
>>> result
[{'emo': 'male_sad', 'score': 91.0}, {'emo': 'male_fear', 'score': 4.8}]