model_xls.py

"""Meta's XLS-R based speaker embedding.
- feature dimension: 768
- source: https://huggingface.co/facebook/wav2vec2-large-xlsr-53
"""
from typing import Optional

import torch
import librosa
import numpy as np
from transformers import AutoFeatureExtractor, AutoModelForPreTraining


class XLSRSE:
    def __init__(self):
        self.processor = AutoFeatureExtractor.from_pretrained("facebook/wav2vec2-large-xlsr-53")
        self.model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-large-xlsr-53")
        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
        self.model.to(self.device)
        self.model.eval()

    def get_speaker_embedding(self, wav: np.ndarray, sampling_rate: Optional[int] = None) -> np.ndarray:
        # audio file is decoded on the fly
        if sampling_rate != self.processor.sampling_rate:
            wav = librosa.resample(wav, orig_sr=sampling_rate, target_sr=self.processor.sampling_rate)
        inputs = self.processor(wav, sampling_rate=self.processor.sampling_rate, return_tensors="pt")
        with torch.no_grad():
            outputs = self.model(**{k: v.to(self.device) for k, v in inputs.items()})
        return outputs.projected_states.mean(1).cpu().numpy()[0]
        # return outputs.projected_quantized_states.mean(1).cpu().numpy()[0]