init

experiment_cache/cluster/xlsr_se.expresso.speaker_id.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2342672f92b65d5671dfb487b78bc5e9cdba199026fa97bcadd778a64a80a210
-size 297705

experiment_cache/cluster/xlsr_se.expresso.style.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ffbf6f8cae2bae0a5175f9f87f4043f34ac672325a96410c472ec54518cec974
-size 331118

experiment_cache/cluster/xlsr_se.voxceleb1-test-split.speaker_id.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cc2b92ab66dd62bda0ccd046ec27ea1c738acfe718adbc748b289f6cab1dc630
-size 184308

experiment_cache/embeddings/xlsr_se.expresso.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c4c78cadd070fab6f469defae141507f0307dbb4f59990757c533631fd6addb8
-size 192055814

experiment_cache/embeddings/xlsr_se.voxceleb1-test-split.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5e34ad588ede7c38ac3591c97e1fc9a8057ca46f61a83d1ad2806601acc775e8
-size 80340078

experiment_cache/tsne/xlsr_se.expresso.speaker_id.npy

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3eb6fcb69c1efb925dabce36b1aefa24f4a87a66029336257d64300590646ba6
-size 93048

experiment_cache/tsne/xlsr_se.expresso.style.npy

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3eb6fcb69c1efb925dabce36b1aefa24f4a87a66029336257d64300590646ba6
-size 93048

experiment_cache/tsne/xlsr_se.voxceleb1-test-split.speaker_id.npy

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e165cffb996e8df023cb1c2a0f4d7173a89b62acdb8b65b9ee33406855248322
-size 39120

experiment_speaker_verification.py

@@ -14,11 +14,13 @@ from sklearn.manifold import TSNE
 import pandas as pd
 from datasets import load_dataset
 
-from model_meta_voice import MetaVoiceSE
-from model_pyannote_embedding import PyannoteSE
-from model_w2v_bert import W2VBertSE
-from model_clap import ClapSE, ClapGeneralSE
-from model_xls import XLSRSE
+from model_meta_voice import MetaVoiceEmbedding
+from model_pyannote_embedding import PyannoteEmbedding
+from model_w2v_bert import W2VBERTEmbedding
+from model_clap import CLAPEmbedding, CLAPGeneralEmbedding
+from model_xls import XLSREmbedding
+from model_hubert import HuBERTXLEmbedding, HuBERTLargeEmbedding
+
 
 
 def get_embedding(model_class, model_name: str, dataset_name: str, data_split: str):
@@ -116,19 +118,23 @@ def analyze_embedding(model_name: str, dataset_name: str, n_shot: int = 5, n_cro
 
 
 if __name__ == '__main__':
-    # get_embedding(MetaVoiceSE, "meta_voice_se", "asahi417/voxceleb1-test-split", "test")
-    # get_embedding(PyannoteSE, "pyannote_se", "asahi417/voxceleb1-test-split", "test")
-    # get_embedding(W2VBertSE, "w2v_bert_se", "asahi417/voxceleb1-test-split", "test")
-    # get_embedding(ClapSE, "clap_se", "asahi417/voxceleb1-test-split", "test")
-    # get_embedding(ClapGeneralSE, "clap_general_se", "asahi417/voxceleb1-test-split", "test")
-    get_embedding(XLSRSE, "xlsr_se", "asahi417/voxceleb1-test-split", "test")
-
-    # get_embedding(MetaVoiceSE, "meta_voice_se", "ylacombe/expresso", "train")
-    # get_embedding(PyannoteSE, "pyannote_se", "ylacombe/expresso", "train")
-    # get_embedding(W2VBertSE, "w2v_bert_se", "ylacombe/expresso", "train")
-    # get_embedding(ClapSE, "clap_se", "ylacombe/expresso", "train")
-    # get_embedding(ClapGeneralSE, "clap_general_se", "ylacombe/expresso", "train")
-    get_embedding(XLSRSE, "xlsr_se", "ylacombe/expresso", "train")
+    # get_embedding(MetaVoiceEmbedding, "meta_voice_se", "asahi417/voxceleb1-test-split", "test")
+    # get_embedding(PyannoteEmbedding, "pyannote_se", "asahi417/voxceleb1-test-split", "test")
+    # get_embedding(W2VBERTEmbedding, "w2v_bert_se", "asahi417/voxceleb1-test-split", "test")
+    # get_embedding(CLAPEmbedding, "clap_se", "asahi417/voxceleb1-test-split", "test")
+    # get_embedding(CLAPGeneralEmbedding, "clap_general_se", "asahi417/voxceleb1-test-split", "test")
+    get_embedding(XLSREmbedding, "xlsr_se", "asahi417/voxceleb1-test-split", "test")
+    get_embedding(HuBERTLargeEmbedding, "hubert_large_se", "asahi417/voxceleb1-test-split", "test")
+    get_embedding(HuBERTXLEmbedding, "hubert_xl_se", "asahi417/voxceleb1-test-split", "test")
+
+    # get_embedding(MetaVoiceEmbedding, "meta_voice_se", "ylacombe/expresso", "train")
+    # get_embedding(PyannoteEmbedding, "pyannote_se", "ylacombe/expresso", "train")
+    # get_embedding(W2VBERTEmbedding, "w2v_bert_se", "ylacombe/expresso", "train")
+    # get_embedding(CLAPEmbedding, "clap_se", "ylacombe/expresso", "train")
+    # get_embedding(CLAPGeneralEmbedding, "clap_general_se", "ylacombe/expresso", "train")
+    get_embedding(XLSREmbedding, "xlsr_se", "ylacombe/expresso", "train")
+    get_embedding(HuBERTLargeEmbedding, "hubert_large_se", "ylacombe/expresso", "train")
+    get_embedding(HuBERTXLEmbedding, "hubert_xl_se", "ylacombe/expresso", "train")
 
     # cluster_embedding("meta_voice_se", "asahi417/voxceleb1-test-split", "speaker_id")
     # cluster_embedding("pyannote_se", "asahi417/voxceleb1-test-split", "speaker_id")
@@ -136,20 +142,26 @@ if __name__ == '__main__':
     # cluster_embedding("clap_se", "asahi417/voxceleb1-test-split", "speaker_id")
     # cluster_embedding("clap_general_se", "asahi417/voxceleb1-test-split", "speaker_id")
     cluster_embedding("xlsr_se", "asahi417/voxceleb1-test-split", "speaker_id")
-    #
+    cluster_embedding("hubert_large_se", "asahi417/voxceleb1-test-split", "speaker_id")
+    cluster_embedding("hubert_xl_se", "asahi417/voxceleb1-test-split", "speaker_id")
+
     # cluster_embedding("meta_voice_se", "ylacombe/expresso", "speaker_id")
     # cluster_embedding("pyannote_se", "ylacombe/expresso", "speaker_id")
     # cluster_embedding("w2v_bert_se", "ylacombe/expresso", "speaker_id")
     # cluster_embedding("clap_se", "ylacombe/expresso", "speaker_id")
     # cluster_embedding("clap_general_se", "ylacombe/expresso", "speaker_id")
     cluster_embedding("xlsr_se", "ylacombe/expresso", "speaker_id")
-    #
+    cluster_embedding("hubert_large_se", "ylacombe/expresso", "speaker_id")
+    cluster_embedding("hubert_xl_se", "ylacombe/expresso", "speaker_id")
+
     # cluster_embedding("meta_voice_se", "ylacombe/expresso", "style")
     # cluster_embedding("pyannote_se", "ylacombe/expresso", "style")
     # cluster_embedding("w2v_bert_se", "ylacombe/expresso", "style")
     # cluster_embedding("clap_se", "ylacombe/expresso", "style")
     # cluster_embedding("clap_general_se", "ylacombe/expresso", "style")
     cluster_embedding("xlsr_se", "ylacombe/expresso", "style")
+    cluster_embedding("hubert_large_se", "ylacombe/expresso", "style")
+    cluster_embedding("hubert_xl_se", "ylacombe/expresso", "style")
 
 
 

model_clap.py

@@ -10,7 +10,7 @@ import numpy as np
 from transformers import ClapModel, ClapProcessor
 
 
-class ClapSE:
+class CLAPEmbedding:
     def __init__(self, ckpt: str = "laion/larger_clap_music_and_speech"):
         self.model = ClapModel.from_pretrained(ckpt)
         self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
@@ -29,7 +29,7 @@ class ClapSE:
         return outputs.cpu().numpy()[0]
 
 
-class ClapGeneralSE(ClapSE):
+class CLAPGeneralEmbedding(CLAPEmbedding):
 
     def __init__(self):
         super().__init__(ckpt="laion/larger_clap_general")

model_hubert.py

@@ -0,0 +1,34 @@
+"""Meta's HuBERT based speaker embedding.
+- feature dimension: 1024
+- source: https://huggingface.co/facebook/hubert-large-ll60k
+"""
+from typing import Optional
+
+import torch
+import librosa
+import numpy as np
+from transformers import AutoFeatureExtractor, AutoModel
+
+
+class HuBERTXLEmbedding:
+
+    def __init__(self, ckpt: str = "facebook/hubert-xlarge-ll60k"):
+        self.processor = AutoFeatureExtractor.from_pretrained(ckpt)
+        self.model = AutoModel.from_pretrained(ckpt)
+        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.last_hidden_state.mean(1).cpu().numpy()[0]
+
+
+class HuBERTLargeEmbedding(HuBERTXLEmbedding):
+    def __init__(self):
+        super().__init__("facebook/hubert-large-ll60k")

model_meta_voice.py

@@ -25,7 +25,7 @@ def wget(url: str, output_file: Optional[str] = None):
         raise ValueError(f"failed to download {url}")
 
 
-class MetaVoiceSE(nn.Module):
+class MetaVoiceEmbedding(nn.Module):
 
     mel_window_length = 25
     mel_window_step = 10

model_pyannote_embedding.py

@@ -11,7 +11,7 @@ from pyannote.audio import Inference
 from pyannote.audio.core.inference import fix_reproducibility, map_with_specifications
 
 
-class PyannoteSE:
+class PyannoteEmbedding:
 
     def __init__(self):
         model = Model.from_pretrained("pyannote/embedding")

model_w2v_bert.py

@@ -10,7 +10,7 @@ import numpy as np
 from transformers import Wav2Vec2BertModel, AutoFeatureExtractor
 
 
-class W2VBertSE:
+class W2VBERTEmbedding:
     def __init__(self):
         self.processor = AutoFeatureExtractor.from_pretrained("facebook/w2v-bert-2.0")
         self.model = Wav2Vec2BertModel.from_pretrained("facebook/w2v-bert-2.0")

model_xls.py

@@ -1,6 +1,7 @@
 """Meta's XLS-R based speaker embedding.
 - feature dimension: 768
 - source: https://huggingface.co/facebook/wav2vec2-large-xlsr-53
+https://huggingface.co/docs/transformers/en/model_doc/wav2vec2#transformers.models.wav2vec2.modeling_wav2vec2.Wav2Vec2ForPreTrainingOutput
 """
 from typing import Optional
 
@@ -10,7 +11,7 @@ import numpy as np
 from transformers import AutoFeatureExtractor, AutoModelForPreTraining
 
 
-class XLSRSE:
+class XLSREmbedding:
     def __init__(self):
         self.processor = AutoFeatureExtractor.from_pretrained("facebook/wav2vec2-large-xlsr-53")
         self.model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-large-xlsr-53")
@@ -26,4 +27,3 @@ class XLSRSE:
         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]

test.py

@@ -1,36 +1,39 @@
 import librosa
-from model_clap import ClapSE
-from model_meta_voice import MetaVoiceSE
-from model_pyannote_embedding import PyannoteSE
-from model_w2v_bert import W2VBertSE
-from model_xls import XLSRSE
+from model_clap import CLAPEmbedding
+from model_meta_voice import MetaVoiceEmbedding
+from model_pyannote_embedding import PyannoteEmbedding
+from model_w2v_bert import W2VBERTEmbedding
+from model_xls import XLSREmbedding
+from model_hubert import HuBERTXLEmbedding
 
 
 def test():
     wav, sr = librosa.load("sample.wav")
     print("XLS-R")
-    model = XLSRSE()
-    v = model.get_speaker_embedding(wav, sr)
-    print(v.shape)
-    model = ClapSE()
+    model = XLSREmbedding()
     v = model.get_speaker_embedding(wav, sr)
     print(v.shape)
     print("CLAP")
-    model = ClapSE()
+    model = CLAPEmbedding()
     v = model.get_speaker_embedding(wav, sr)
     print(v.shape)
     print("MetaVoiceSE")
-    model = MetaVoiceSE()
+    model = MetaVoiceEmbedding()
     v = model.get_speaker_embedding(wav, sr)
     print(v.shape)
     print("PyannoteSE")
-    model = PyannoteSE()
+    model = PyannoteEmbedding()
     v = model.get_speaker_embedding(wav, sr)
     print(v.shape)
     print("W2VBertSE")
-    model = W2VBertSE()
+    model = W2VBERTEmbedding()
     v = model.get_speaker_embedding(wav, sr)
     print(v.shape)
+    print("huBERT")
+    model = HuBERTXLEmbedding()
+    v = model.get_speaker_embedding(wav, sr)
+    print(v.shape)
+
 
 
 if __name__ == '__main__':