compute embeddings and create speakers.json

TTS/bin/compute_embeddings.py

@@ -9,9 +9,11 @@ import torch
 from TTS.speaker_encoder.model import SpeakerEncoder
 from TTS.utils.audio import AudioProcessor
 from TTS.utils.io import load_config
+from TTS.utils.io import save_speaker_mapping
+from TTS.tts.datasets.preprocess import load_meta_data
 
 parser = argparse.ArgumentParser(
-    description='Compute embedding vectors for each wav file in a dataset. ')
+    description='Compute embedding vectors for each wav file in a dataset. If "target_dataset" is defined, it generates "speakers.json" necessary for training a multi-speaker model.')
 parser.add_argument(
     'model_path',
     type=str,
@@ -29,6 +31,12 @@ parser.add_argument(
     'output_path',
     type=str,
     help='path for training outputs.')
+parser.add_argument(
+    '--target_dataset',
+    type=str,
+    default='',
+    help='Target dataset to pick a processor from TTS.tts.dataset.preprocess. Necessary to create a speakers.json file.'
+)
 parser.add_argument(
     '--use_cuda', type=bool, help='flag to set cuda.', default=False
 )
@@ -45,44 +53,76 @@ data_path = args.data_path
 split_ext = os.path.splitext(data_path)
 sep = args.separator
 
-if len(split_ext) > 0 and split_ext[1].lower() == '.csv':
+if args.target_dataset != '':
-    # Parse CSV
+    # if target dataset is defined
-    print(f'CSV file: {data_path}')
+    dataset_config = [
-    with open(data_path) as f:
+        {
-        wav_path = os.path.join(os.path.dirname(data_path), 'wavs')
+            "name": args.target_dataset,
-        wav_files = []
+            "path": args.data_path,
-        print(f'Separator is: {sep}')
+            "meta_file_train": None,
-        for line in f:
+            "meta_file_val": None
-            components = line.split(sep)
+        },
-            if len(components) != 2:
+    ]
-                print("Invalid line")
+    wav_files, _ = load_meta_data(dataset_config, eval_split=False)
-                continue
+    output_files = [wav_file[1].replace(data_path, args.output_path).replace(
-            wav_file = os.path.join(wav_path, components[0] + '.wav')
+        '.wav', '.npy') for wav_file in wav_files]
-            #print(f'wav_file: {wav_file}')
-            if os.path.exists(wav_file):
-                wav_files.append(wav_file)
-    print(f'Count of wavs imported: {len(wav_files)}')
 else:
-    # Parse all wav files in data_path
+    # if target dataset is not defined
-    wav_path = data_path
+    if len(split_ext) > 0 and split_ext[1].lower() == '.csv':
-    wav_files = glob.glob(data_path + '/**/*.wav', recursive=True)
+        # Parse CSV
+        print(f'CSV file: {data_path}')
+        with open(data_path) as f:
+            wav_path = os.path.join(os.path.dirname(data_path), 'wavs')
+            wav_files = []
+            print(f'Separator is: {sep}')
+            for line in f:
+                components = line.split(sep)
+                if len(components) != 2:
+                    print("Invalid line")
+                    continue
+                wav_file = os.path.join(wav_path, components[0] + '.wav')
+                #print(f'wav_file: {wav_file}')
+                if os.path.exists(wav_file):
+                    wav_files.append(wav_file)
+        print(f'Count of wavs imported: {len(wav_files)}')
+    else:
+        # Parse all wav files in data_path
+        wav_files = glob.glob(data_path + '/**/*.wav', recursive=True)
 
-output_files = [wav_file.replace(wav_path, args.output_path).replace(
+        output_files = [wav_file.replace(data_path, args.output_path).replace(
-    '.wav', '.npy') for wav_file in wav_files]
+            '.wav', '.npy') for wav_file in wav_files]
 
 for output_file in output_files:
     os.makedirs(os.path.dirname(output_file), exist_ok=True)
 
+# define Encoder model
 model = SpeakerEncoder(**c.model)
 model.load_state_dict(torch.load(args.model_path)['model'])
 model.eval()
 if args.use_cuda:
     model.cuda()
 
+# compute speaker embeddings
+speaker_mapping = {}
 for idx, wav_file in enumerate(tqdm(wav_files)):
+    if isinstance(wav_file, list):
+        speaker_name = wav_file[2]
+        wav_file = wav_file[1]
     mel_spec = ap.melspectrogram(ap.load_wav(wav_file, sr=ap.sample_rate)).T
     mel_spec = torch.FloatTensor(mel_spec[None, :, :])
     if args.use_cuda:
         mel_spec = mel_spec.cuda()
     embedd = model.compute_embedding(mel_spec)
     np.save(output_files[idx], embedd.detach().cpu().numpy())
+
+    if args.target_dataset != '':
+        # create speaker_mapping if target dataset is defined
+        wav_file_name = os.path.basename(wav_file)
+        speaker_mapping[wav_file_name] = {}
+        speaker_mapping[wav_file_name]['name'] = speaker_name
+        speaker_mapping[wav_file_name]['embedding'] = embedd.detach().cpu().numpy()
+
+if args.target_dataset != '':
+    # save speaker_mapping if target dataset is defined
+    mapping_file_path = os.path.join(args.output_path, 'speakers.json')
+    save_speaker_mapping(mapping_file_path, speaker_mapping)