Update app.py

app.py

@@ -1,12 +1,12 @@
 import gradio as gr
 import numpy as np
 import librosa
-import time
 import requests
 from io import BytesIO
 from PIL import Image
 import os
 from tensorflow.keras.models import load_model
+from faster_whisper import WhisperModel
 
 # Load the emotion prediction model
 def load_emotion_model(model_path):
@@ -20,20 +20,15 @@ def load_emotion_model(model_path):
 model_path = 'mymodel_SER_LSTM_RAVDESS.h5'
 model = load_emotion_model(model_path)
 
-#####
-
-from faster_whisper import WhisperModel
-
-
+# Initialize WhisperModel
 model_size = "small"
 model2 = WhisperModel(model_size, device="cpu", compute_type="int8")
 
+# Function to transcribe audio
 def transcribe(wav_filepath):
-    segments, _ = model2.transcribe(wav_filepath, beam_size=5)    
+    segments, _ = model2.transcribe(wav_filepath, beam_size=5)
     return "".join([segment.text for segment in segments])
 
-#########
-    
 # Function to extract MFCC features from audio
 def extract_mfcc(wav_file_name):
     try:
@@ -64,24 +59,13 @@ def predict_emotion_from_audio(wav_filepath):
 
 api_key = os.getenv("DeepAI_api_key")
 
-
-# Predict emotion from audio
-def get_predictions(audio_input):
-    emotion_prediction = predict_emotion_from_audio(audio_input)
-    # Generate image here or call a separate function
-    image = generate_image(api_key, emotion_prediction)
-    return emotion_prediction, image
-
-
-# Define a function to generate an image using DeepAI Text to Image API
+# Function to generate an image using DeepAI Text to Image API
 def generate_image(api_key, text):
     url = "https://api.deepai.org/api/text2img"
     headers = {'api-key': api_key}
     response = requests.post(
         url,
-        data={
-            'text': text,
-        },
+        data={'text': text},
         headers=headers
     )
     response_data = response.json()
@@ -92,29 +76,18 @@ def generate_image(api_key, text):
         return image
     else:
         return None
-####
 
+# Function to get predictions
+def get_predictions(audio_input):
+    emotion_prediction = predict_emotion_from_audio(audio_input)
+    image = generate_image(api_key, emotion_prediction)
+    return emotion_prediction, image
 
 # Create the Gradio interface
-with gr.Blocks() as interface:
-    gr.Markdown("Emotional Machines test: Load or Record an audio file to speech emotion analysis")
-
-    with gr.Tabs():
-        with gr.Tab("Acoustic and Semantic Predictions"):
-            with gr.Row():
-                input_audio = gr.Audio(label="Input Audio", type="filepath")
-                submit_button = gr.Button("Submit")
-            output_label = [gr.Label("Prediction"), gr.Image(type='pil')]  # Use a single Label instead of a list
-
-            def submit_callback(wav_filepath):
-                emotion_prediction, _ = get_predictions(wav_filepath)  # Extract only the emotion prediction
-                return emotion_prediction
-
-# Set the function to be called when the button is clicked
-submit_button.click(submit_callback, inputs=input_audio, outputs=output_label[0])
-
-interface.launch()
-        
-        
-
+interface = gr.Interface(
+    fn=get_predictions,
+    inputs=gr.Audio(label="Input Audio", type="file"),
+    outputs=[gr.Label("Prediction"), gr.Image(type='pil')]
+)
 
+interface.launch()