Merge branch 'dev-emo' into master

emo_gen.py

@@ -0,0 +1,169 @@
+import torch
+import torch.nn as nn
+from torch.utils.data import Dataset
+from torch.utils.data import DataLoader
+from transformers import Wav2Vec2Processor
+from transformers.models.wav2vec2.modeling_wav2vec2 import (
+    Wav2Vec2Model,
+    Wav2Vec2PreTrainedModel,
+)
+import librosa
+import numpy as np
+import argparse
+from config import config
+import utils
+import os
+from tqdm import tqdm
+
+
+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)
+
+
+model_name = "./emotional/wav2vec2-large-robust-12-ft-emotion-msp-dim"
+processor = Wav2Vec2Processor.from_pretrained(model_name)
+model = EmotionModel.from_pretrained(model_name)
+
+
+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)
+    device = config.bert_gen_config.device
+    return process_func(
+        np.expand_dims(wav, 0).astype(np.float),
+        sr,
+        model,
+        processor,
+        device,
+        embeddings=True,
+    ).squeeze(0)
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "-c", "--config", type=str, default=config.bert_gen_config.config_path
+    )
+    parser.add_argument(
+        "--num_processes", type=int, default=config.bert_gen_config.num_processes
+    )
+    args, _ = parser.parse_known_args()
+    config_path = args.config
+    hps = utils.get_hparams_from_file(config_path)
+
+    device = config.bert_gen_config.device
+
+    model_name = "./emotional/wav2vec2-large-robust-12-ft-emotion-msp-dim"
+    processor = (
+        Wav2Vec2Processor.from_pretrained(model_name)
+        if processor is None
+        else processor
+    )
+    model = (
+        EmotionModel.from_pretrained(model_name).to(device)
+        if model is None
+        else model.to(device)
+    )
+
+    lines = []
+    with open(hps.data.training_files, encoding="utf-8") as f:
+        lines.extend(f.readlines())
+
+    with open(hps.data.validation_files, encoding="utf-8") as f:
+        lines.extend(f.readlines())
+
+    wavnames = [line.split("|")[0] for line in lines]
+    dataset = AudioDataset(wavnames, 16000, processor)
+    data_loader = DataLoader(dataset, batch_size=1, shuffle=False, num_workers=16)
+
+    with torch.no_grad():
+        for i, data in tqdm(enumerate(data_loader), total=len(data_loader)):
+            wavname = wavnames[i]
+            emo_path = wavname.replace(".wav", ".emo.npy")
+            if os.path.exists(emo_path):
+                continue
+            emb = model(data.to(device))[0].detach().cpu().numpy()
+            np.save(emo_path, emb)
+
+    print("Emo vec 生成完毕!")

models.py

@@ -10,6 +10,7 @@ import monotonic_align
 
 from torch.nn import Conv1d, ConvTranspose1d, Conv2d
 from torch.nn.utils import weight_norm, remove_weight_norm, spectral_norm
+from vector_quantize_pytorch import VectorQuantize
 
 from commons import init_weights, get_padding
 from text import symbols, num_tones, num_languages
@@ -321,6 +322,7 @@ class TextEncoder(nn.Module):
         n_layers,
         kernel_size,
         p_dropout,
+        n_speakers,
         gin_channels=0,
     ):
         super().__init__()
@@ -342,6 +344,18 @@ class TextEncoder(nn.Module):
         self.bert_proj = nn.Conv1d(1024, hidden_channels, 1)
         self.ja_bert_proj = nn.Conv1d(1024, hidden_channels, 1)
         self.en_bert_proj = nn.Conv1d(1024, hidden_channels, 1)
+        self.emo_proj = nn.Linear(1024, 1024)
+        self.emo_quantizer = [
+            VectorQuantize(
+                dim=1024,
+                codebook_size=5,
+                decay=0.8,
+                commitment_weight=1.0,
+                learnable_codebook=True,
+                ema_update=False,
+            )
+        ] * n_speakers
+        self.emo_q_proj = nn.Linear(1024, hidden_channels)
 
         self.encoder = attentions.Encoder(
             hidden_channels,
@@ -354,10 +368,33 @@ class TextEncoder(nn.Module):
         )
         self.proj = nn.Conv1d(hidden_channels, out_channels * 2, 1)
 
-    def forward(self, x, x_lengths, tone, language, bert, ja_bert, en_bert, g=None):
+    def forward(
+        self, x, x_lengths, tone, language, bert, ja_bert, en_bert, emo, sid, g=None
+    ):
+        sid = sid.cpu()
         bert_emb = self.bert_proj(bert).transpose(1, 2)
         ja_bert_emb = self.ja_bert_proj(ja_bert).transpose(1, 2)
         en_bert_emb = self.en_bert_proj(en_bert).transpose(1, 2)
+        if emo.size(-1) == 1024:
+            emo_emb = self.emo_proj(emo.unsqueeze(1))
+            emo_commit_loss = torch.zeros(1)
+            emo_emb_ = []
+            for i in range(emo_emb.size(0)):
+                temp_emo_emb, _, temp_emo_commit_loss = self.emo_quantizer[sid[i]](
+                    emo_emb[i].unsqueeze(0).cpu()
+                )
+                emo_commit_loss += temp_emo_commit_loss
+                emo_emb_.append(temp_emo_emb)
+            emo_emb = torch.cat(emo_emb_, dim=0).to(emo_emb.device)
+            emo_commit_loss = emo_commit_loss.to(emo_emb.device)
+        else:
+            emo_emb = (
+                self.emo_quantizer[sid[0]]
+                .get_output_from_indices(emo.to(torch.int).cpu())
+                .unsqueeze(0)
+                .to(emo.device)
+            )
+            emo_commit_loss = torch.zeros(1)
         x = (
             self.emb(x)
             + self.tone_emb(tone)
@@ -365,6 +402,7 @@ class TextEncoder(nn.Module):
             + bert_emb
             + ja_bert_emb
             + en_bert_emb
+            + self.emo_q_proj(emo_emb)
         ) * math.sqrt(
             self.hidden_channels
         )  # [b, t, h]
@@ -377,7 +415,7 @@ class TextEncoder(nn.Module):
         stats = self.proj(x) * x_mask
 
         m, logs = torch.split(stats, self.out_channels, dim=1)
-        return x, m, logs, x_mask
+        return x, m, logs, x_mask, emo_commit_loss
 
 
 class ResidualCouplingBlock(nn.Module):
@@ -810,6 +848,7 @@ class SynthesizerTrn(nn.Module):
             n_layers,
             kernel_size,
             p_dropout,
+            self.n_speakers,
             gin_channels=self.enc_gin_channels,
         )
         self.dec = Generator(
@@ -860,7 +899,7 @@ class SynthesizerTrn(nn.Module):
             hidden_channels, 256, 3, 0.5, gin_channels=gin_channels
         )
 
-        if n_speakers >= 1:
+        if n_speakers > =1:
             self.emb_g = nn.Embedding(n_speakers, gin_channels)
         else:
             self.ref_enc = ReferenceEncoder(spec_channels, gin_channels)
@@ -877,13 +916,14 @@ class SynthesizerTrn(nn.Module):
         bert,
         ja_bert,
         en_bert,
+        emo=None,
     ):
         if self.n_speakers > 0:
             g = self.emb_g(sid).unsqueeze(-1)  # [b, h, 1]
         else:
             g = self.ref_enc(y.transpose(1, 2)).unsqueeze(-1)
-        x, m_p, logs_p, x_mask = self.enc_p(
+        x, m_p, logs_p, x_mask, loss_commit = self.enc_p(
-            x, x_lengths, tone, language, bert, ja_bert, en_bert, g=g
+            x, x_lengths, tone, language, bert, ja_bert, en_bert, emo, sid, g=g
         )
         z, m_q, logs_q, y_mask = self.enc_q(y, y_lengths, g=g)
         z_p = self.flow(z, y_mask, g=g)
@@ -949,6 +989,7 @@ class SynthesizerTrn(nn.Module):
             y_mask,
             (z, z_p, m_p, logs_p, m_q, logs_q),
             (x, logw, logw_),
+            loss_commit,
         )
 
     def infer(
@@ -961,6 +1002,7 @@ class SynthesizerTrn(nn.Module):
         bert,
         ja_bert,
         en_bert,
+        emo=None,
         noise_scale=0.667,
         length_scale=1,
         noise_scale_w=0.8,
@@ -974,8 +1016,8 @@ class SynthesizerTrn(nn.Module):
             g = self.emb_g(sid).unsqueeze(-1)  # [b, h, 1]
         else:
             g = self.ref_enc(y.transpose(1, 2)).unsqueeze(-1)
-        x, m_p, logs_p, x_mask = self.enc_p(
+        x, m_p, logs_p, x_mask, _ = self.enc_p(
-            x, x_lengths, tone, language, bert, ja_bert, en_bert, g=g
+            x, x_lengths, tone, language, bert, ja_bert, en_bert, emo, sid, g=g
         )
         logw = self.sdp(x, x_mask, g=g, reverse=True, noise_scale=noise_scale_w) * (
             sdp_ratio

train_ms.py

@@ -340,6 +340,7 @@ def train_and_evaluate(
         bert,
         ja_bert,
         en_bert,
+        emo,
     ) in tqdm(enumerate(train_loader)):
         if net_g.module.use_noise_scaled_mas:
             current_mas_noise_scale = (
@@ -362,6 +363,7 @@ def train_and_evaluate(
         bert = bert.cuda(rank, non_blocking=True)
         ja_bert = ja_bert.cuda(rank, non_blocking=True)
         en_bert = en_bert.cuda(rank, non_blocking=True)
+        emo = emo.cuda(rank, non_blocking=True)
 
         with autocast(enabled=hps.train.fp16_run):
             (
@@ -373,6 +375,7 @@ def train_and_evaluate(
                 z_mask,
                 (z, z_p, m_p, logs_p, m_q, logs_q),
                 (hidden_x, logw, logw_),
+                loss_commit,
             ) = net_g(
                 x,
                 x_lengths,
@@ -384,6 +387,7 @@ def train_and_evaluate(
                 bert,
                 ja_bert,
                 en_bert,
+                emo,
             )
             mel = spec_to_mel_torch(
                 spec,
@@ -454,7 +458,9 @@ def train_and_evaluate(
 
                 loss_fm = feature_loss(fmap_r, fmap_g)
                 loss_gen, losses_gen = generator_loss(y_d_hat_g)
-                loss_gen_all = loss_gen + loss_fm + loss_mel + loss_dur + loss_kl
+                loss_gen_all = (
+                    loss_gen + loss_fm + loss_mel + loss_dur + loss_kl + loss_commit
+                )
                 if net_dur_disc is not None:
                     loss_dur_gen, losses_dur_gen = generator_loss(y_dur_hat_g)
                     loss_gen_all += loss_dur_gen
@@ -579,6 +585,7 @@ def evaluate(hps, generator, eval_loader, writer_eval):
             bert,
             ja_bert,
             en_bert,
+            emo,
         ) in enumerate(eval_loader):
             x, x_lengths = x.cuda(), x_lengths.cuda()
             spec, spec_lengths = spec.cuda(), spec_lengths.cuda()