[demo] update wespeakerruntime version

app.py

@@ -15,7 +15,7 @@
 
 import gradio as gr
 import wespeakerruntime as wespeaker
-from sklearn.metrics.pairwise import cosine_similarity
+# from sklearn.metrics.pairwise import cosine_similarity
 
 STYLE = """
 <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bootstrap@5.1.3/dist/css/bootstrap.min.css" integrity="sha256-YvdLHPgkqJ8DVUxjjnGVlMMJtNimJ6dYkowFFvp4kKs=" crossorigin="anonymous">
@@ -46,8 +46,8 @@ OUTPUT_ERROR = (STYLE + """
     </div>
 """)
 
-vox_model = wespeaker.Inference('pre_model/voxceleb_resnet34.onnx')
-cnc_model = wespeaker.Inference('pre_model/cnceleb_resnet34.onnx')
+vox_model = wespeaker.Speaker('pre_model/voxceleb_resnet34.onnx')
+cnc_model = wespeaker.Speaker('pre_model/cnceleb_resnet34.onnx')
 
 
 def speaker_verification(audio_path1, audio_path2, lang='CN'):
@@ -61,10 +61,9 @@ def speaker_verification(audio_path1, audio_path2, lang='CN'):
     else:
         output = OUTPUT_ERROR.format('Please select a language')
         return output
-    emb1 = model.extract_embedding_wav(audio_path1)
-    emb2 = model.extract_embedding_wav(audio_path2)
-    cos_score = cosine_similarity(emb1.reshape(1, -1), emb2.reshape(1,
-                                                                    -1))[0][0]
+    emb1 = model.extract_embedding(audio_path1)
+    emb2 = model.extract_embedding(audio_path2)
+    cos_score = model.compute_cosine_score(emb1[0], emb2[0])
     cos_score = (cos_score + 1) / 2.0
 
     if cos_score >= 0.70: