File size: 3,083 Bytes
1cf1e13
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
import librosa
import numpy as np
import torch
import torch.nn as nn
from torch.utils.data import Dataset
from torch.utils.data import Dataset
from transformers import Wav2Vec2Processor
from transformers.models.wav2vec2.modeling_wav2vec2 import (
    Wav2Vec2Model,
    Wav2Vec2PreTrainedModel,
)

from config import config


class RegressionHead(nn.Module):
    r"""Classification head."""

    def __init__(self, config):
        super().__init__()

        self.dense = nn.Linear(config.hidden_size, config.hidden_size)
        self.dropout = nn.Dropout(config.final_dropout)
        self.out_proj = nn.Linear(config.hidden_size, config.num_labels)

    def forward(self, features, **kwargs):
        x = features
        x = self.dropout(x)
        x = self.dense(x)
        x = torch.tanh(x)
        x = self.dropout(x)
        x = self.out_proj(x)

        return x


class EmotionModel(Wav2Vec2PreTrainedModel):
    r"""Speech emotion classifier."""

    def __init__(self, config):
        super().__init__(config)

        self.config = config
        self.wav2vec2 = Wav2Vec2Model(config)
        self.classifier = RegressionHead(config)
        self.init_weights()

    def forward(
        self,
        input_values,
    ):
        outputs = self.wav2vec2(input_values)
        hidden_states = outputs[0]
        hidden_states = torch.mean(hidden_states, dim=1)
        logits = self.classifier(hidden_states)

        return hidden_states, logits


class AudioDataset(Dataset):
    def __init__(self, list_of_wav_files, sr, processor):
        self.list_of_wav_files = list_of_wav_files
        self.processor = processor
        self.sr = sr

    def __len__(self):
        return len(self.list_of_wav_files)

    def __getitem__(self, idx):
        wav_file = self.list_of_wav_files[idx]
        audio_data, _ = librosa.load(wav_file, sr=self.sr)
        processed_data = self.processor(audio_data, sampling_rate=self.sr)[
            "input_values"
        ][0]
        return torch.from_numpy(processed_data)


device = config.emo_gen_config.device
model_name = "./emotional/wav2vec2-large-robust-12-ft-emotion-msp-dim"
processor = Wav2Vec2Processor.from_pretrained(model_name)
model = EmotionModel.from_pretrained(model_name).to(device)


def process_func(
    x: np.ndarray,
    sampling_rate: int,
    model: EmotionModel,
    processor: Wav2Vec2Processor,
    device: str,
    embeddings: bool = False,
) -> np.ndarray:
    r"""Predict emotions or extract embeddings from raw audio signal."""
    model = model.to(device)
    y = processor(x, sampling_rate=sampling_rate)
    y = y["input_values"][0]
    y = torch.from_numpy(y).unsqueeze(0).to(device)

    # run through model
    with torch.no_grad():
        y = model(y)[0 if embeddings else 1]

    # convert to numpy
    y = y.detach().cpu().numpy()

    return y


def get_emo(path):
    wav, sr = librosa.load(path, 16000)
    return process_func(
        np.expand_dims(wav, 0).astype(np.float64),
        sr,
        model,
        processor,
        device,
        embeddings=True,
    ).squeeze(0)