Update app.py

app.py

@@ -1,7 +1,30 @@
-# Set up the Gradio interface
+#Set up the Gradio interface
+
 import gradio as gr
+from transformers import pipeline
+from TTS.api import TTS
+
+# Load pre-trained emotion detection model
+emotion_classifier = pipeline("text-classification", model="bhadresh-savani/distilbert-base-uncased-emotion")
+
+# Load TTS model
+tts_model = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC")
+
+# Emotion-specific settings for pitch and speed
+emotion_settings = {
+    "neutral": {"pitch": 1.0, "speed": 1.0},
+    "joy": {"pitch": 1.3, "speed": 1.2},
+    "sadness": {"pitch": 0.8, "speed": 0.9},
+    "anger": {"pitch": 1.6, "speed": 1.4},
+    "fear": {"pitch": 1.2, "speed": 0.95},
+    "surprise": {"pitch": 1.5, "speed": 1.3},
+    "disgust": {"pitch": 0.9, "speed": 0.95},
+    "shame": {"pitch": 0.8, "speed": 0.85},
+}
 
-def emotion_aware_tts_pipeline_gradio(input_text=None, file_input=None):
+
+# Function to process text or file input and generate audio
+def emotion_aware_tts_pipeline(input_text=None, file_input=None):
     try:
         # Get text from input or file
         if file_input:
@@ -21,11 +44,9 @@ def emotion_aware_tts_pipeline_gradio(input_text=None, file_input=None):
 
             # Generate audio
             audio_path = "output.wav"
-            mel_spectrogram = tts_model.get_mel_spectrogram(input_text)
-            audio = vocoder.decode(mel_spectrogram)
+            tts_model.tts_to_file(text=input_text, file_path=audio_path, speed=speed, pitch=pitch)
+
 
-            # Post-processing: adjust pitch and speed
-            adjust_pitch_and_speed(audio_path, pitch_factor=pitch, speed_factor=speed)
 
             return f"Detected Emotion: {emotion} (Confidence: {confidence:.2f})", audio_path
         else:
@@ -33,9 +54,11 @@ def emotion_aware_tts_pipeline_gradio(input_text=None, file_input=None):
     except Exception as e:
         return f"Error: {str(e)}", None
 
+
+
 # Define Gradio interface
 iface = gr.Interface(
-    fn=emotion_aware_tts_pipeline_gradio,
+    fn=emotion_aware_tts_pipeline,
     inputs=[
         gr.Textbox(label="Input Text", placeholder="Enter text here"),
         gr.File(label="Upload a Text File")
@@ -49,4 +72,4 @@ iface = gr.Interface(
 )
 
 # Launch Gradio interface
-iface.launch()
+iface.launch()