Upload app.py

app.py

@@ -0,0 +1,46 @@
+from speechbrain.pretrained.interfaces import foreign_class
+import gradio as gr
+import nltk
+import librosa
+
+import warnings
+warnings.filterwarnings("ignore")
+
+# Loading the speechbrain emotion detection model
+learner = foreign_class(source="mtauro/wavlm_vrs_ck_iva_k492", 
+                        pymodule_file="custom_interface.py",
+                          classname="CustomEncoderWav2vec2Classifier")
+
+
+# Building prediction function for gradio
+emotion_dict = {
+    'HC': 'No AD detected', 
+    'AD': 'Signs of AD detected',
+
+}
+
+#def predict_emotion(audio):
+#    out_prob, score, index, text_lab = learner.classify_file(audio.name)
+#    return emotion_dict[text_lab[0]]
+
+def detect_ad(audio):
+    out_prob, score, index, text_lab = learner.classify_file(audio.name)
+    return emotion_dict[text_lab[0]]    
+
+# Loading gradio interface
+#inputs = gr.inputs.Audio(label="Input Audio", type="file")
+#input_1 = gr.inputs.Audio(label="Input Audio", type="file")
+##inputs = gr.inputs.Audio(source="microphone", type="filepath", optional=True, label="Please record your voice")
+#outputs = "text"
+#title = "AD Detection"
+#description = "Testing"
+#gr.Interface(detect_ad, inputs, outputs, title=title, description=description).launch()
+#gr.Interface(detect_ad,  input_1, outputs, title=title, description=description).launch()
+
+gr.Interface(detect_ad,
+             inputs = gr.inputs.Audio(source="microphone", type="file", optional=True, label="Please speak for 13-30 seconds"),
+             outputs = gr.outputs.Textbox(label="Output Text"),
+             title="AD Detection",
+             description = "Testing",
+             #examples = [["Test_File1.wav"], ["Test_File2.wav"], ["Test_File3.wav"]], 
+             theme="grass").launch()