Update app.py

app.py

@@ -1,44 +1,120 @@
 import gradio as gr
 import torch
 import torchaudio
+# from torchaudio.sox_effects import apply_effects_file
 from transformers import AutoFeatureExtractor, AutoModelForAudioXVector
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
-def similarity_fn(speaker1, speaker2):
-  if not (speaker1 and speaker2):
-      return gr.Textbox(value='<b style="color:red">ERROR: Please record audio for *both* speakers!</b>')
-
-  wav1, _ = torchaudio.load(speaker1)
-  wav2, _ = torchaudio.load(speaker2)
-
-  feature_extractor = AutoFeatureExtractor.from_pretrained("microsoft/wavlm-base-plus-sv")
-  model = AutoModelForAudioXVector.from_pretrained("microsoft/wavlm-base-plus-sv").to(device)
-  
-  input1 = feature_extractor(wav1.squeeze(0), return_tensors="pt", sampling_rate=16000).input_values.to(device)
-  input2 = feature_extractor(wav2.squeeze(0), return_tensors="pt", sampling_rate=16000).input_values.to(device)
-
-  with torch.no_grad():
-      emb1 = model(input1).embeddings  
-      emb2 = model(input2).embeddings
-  emb1 = torch.nn.functional.normalize(emb1, dim=-1).cpu()
-  emb2 = torch.nn.functional.normalize(emb2, dim=-1).cpu()
-  similarity = torch.nn.CosineSimilarity(dim=-1)(emb1, emb2).numpy()[0]
-
-  if similarity >= 0.8:
-      label = "The speakers are similar"
-      color = "green"
-  else:
-      label = "The speakers are different"
-      color = "red"
-  
-  return gr.Textbox(value=f"<span style='color:{color}'>{label}</span>")
-
-demo = gr.Interface(
-    speaker1=gr.Audio(source="microphone", type="filepath"),
-    speaker2=gr.Audio(source="microphone", type="filepath"),
-    output=gr.Textbox(), 
-    fn=similarity_fn
+STYLE = """
+<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" integrity="sha256-YvdLHPgkqJ8DVUxjjnGVlMMJtNimJ6dYkowFFvp4kKs=" crossorigin="anonymous">
+"""
+OUTPUT_OK = (
+    STYLE
+    + """
+    <div class="container">
+        <div class="row"><h1 style="text-align: center">The speakers are</h1></div>
+        <div class="row"><h1 class="display-1 text-success" style="text-align: center">{:.1f}%</h1></div>
+        <div class="row"><h1 style="text-align: center">similar</h1></div>
+        <div class="row"><h1 class="text-success" style="text-align: center">Welcome, human!</h1></div>
+        <div class="row"><small style="text-align: center">(You must get at least 80% to be considered the same person)</small><div class="row">
+    </div>
+"""
 )
+OUTPUT_FAIL = (
+    STYLE
+    + """
+    <div class="container">
+        <div class="row"><h1 style="text-align: center">The speakers are</h1></div>
+        <div class="row"><h1 class="display-1 text-danger" style="text-align: center">{:.1f}%</h1></div>
+        <div class="row"><h1 style="text-align: center">similar</h1></div>
+        <div class="row"><h1 class="text-danger" style="text-align: center">You shall not pass!</h1></div>
+        <div class="row"><small style="text-align: center">(You must get at least 80% to be considered the same person)</small><div class="row">
+    </div>
+"""
+)
+
+EFFECTS = [
+    ["remix", "-"],
+    ["channels", "1"],
+    ["rate", "16000"],
+    ["gain", "-1.0"],
+    ["silence", "1", "0.1", "0.1%", "-1", "0.1", "0.1%"],
+    ["trim", "0", "10"],
+]
+
+THRESHOLD = 0.80
+
+model_name = "microsoft/wavlm-base-plus-sv"
+feature_extractor = AutoFeatureExtractor.from_pretrained(model_name)
+model = AutoModelForAudioXVector.from_pretrained(model_name).to(device)
+cosine_sim = torch.nn.CosineSimilarity(dim=-1)
+
+
+def similarity_fn(path1, path2):
+    if not (path1 and path2):
+        return '<b style="color:red">ERROR: Please record audio for *both* speakers!</b>'
+
+    # wav1, _ = apply_effects_file(path1, EFFECTS)
+    # wav2, _ = apply_effects_file(path2, EFFECTS)
+    wav1, _ = torchaudio.load(path1)
+    wav2, _ = torchaudio.load(path2)
+    print(wav1.shape, wav2.shape)
+
+    input1 = feature_extractor(wav1.squeeze(0), return_tensors="pt", sampling_rate=16000).input_values.to(device)
+    input2 = feature_extractor(wav2.squeeze(0), return_tensors="pt", sampling_rate=16000).input_values.to(device)
+
+    with torch.no_grad():
+        emb1 = model(input1).embeddings
+        emb2 = model(input2).embeddings
+    emb1 = torch.nn.functional.normalize(emb1, dim=-1).cpu()
+    emb2 = torch.nn.functional.normalize(emb2, dim=-1).cpu()
+    similarity = cosine_sim(emb1, emb2).numpy()[0]
 
-demo.launch()
+    if similarity >= THRESHOLD:
+        output = OUTPUT_OK.format(similarity * 100)
+    else:
+        output = OUTPUT_FAIL.format(similarity * 100)
+
+    return output
+
+
+inputs = [
+    gr.Audio(sources=["microphone", "upload"], type="filepath", label="Speaker #1"),
+    gr.Audio(sources=["microphone", "upload"], type="filepath", label="Speaker #2"),
+]
+output = gr.HTML(label="")
+
+
+description = (
+    "This demo will compare two speech samples and determine if they are from the same speaker. "
+    "Try it with your own voice!"
+)
+article = (
+    "<p style='text-align: center'>"
+    "<a href='https://huggingface.co/microsoft/wavlm-base-plus-sv' target='_blank'>🎙️ Learn more about WavLM</a> | "
+    "<a href='https://arxiv.org/abs/2110.13900' target='_blank'>📚 WavLM paper</a> | "
+    "<a href='https://www.danielpovey.com/files/2018_icassp_xvectors.pdf' target='_blank'>📚 X-Vector paper</a>"
+    "</p>"
+)
+examples = [
+    ["samples/denzel_washington.mp3", "samples/denzel_washington.mp3"],
+    ["samples/heath_ledger_2.mp3", "samples/heath_ledger_3.mp3"],
+    ["samples/heath_ledger_3.mp3", "samples/denzel_washington.mp3"],
+    ["samples/denzel_washington.mp3", "samples/heath_ledger_2.mp3"],
+]
+
+interface = gr.Interface(
+    fn=similarity_fn,
+    inputs=inputs,
+    outputs=output,
+    title="Voice Authentication with WavLM + X-Vectors",
+    description=description,
+    article=article,
+    layout="horizontal",
+    theme="huggingface",
+    allow_flagging=False,
+    live=False,
+    examples=examples,
+)
+interface.launch(enable_queue=True)