Add AI Virtual Therapist application

app.py

@@ -0,0 +1,121 @@
+import gradio as gr
+from fastapi import FastAPI, UploadFile, File, HTTPException
+import librosa
+import openai
+from transformers import pipeline
+import requests
+import os
+from pydantic import BaseModel
+import numpy as np
+
+# Initialize FastAPI
+app = FastAPI()
+
+# Initialize emotion classifier
+text_emotion_classifier = pipeline("text-classification", 
+                                 model="bhadresh-savani/distilbert-base-uncased-emotion",
+                                 device=-1)  # Use CPU
+
+# Environment variables will be set in Hugging Face Spaces
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+ELEVEN_LABS_API_KEY = os.getenv("ELEVEN_LABS_API_KEY")
+VOICE_ID = os.getenv("VOICE_ID", "9BWtsMINqrJLrRacOk9x")
+
+def analyze_text_emotion(text):
+    try:
+        emotion_result = text_emotion_classifier(text)
+        emotion_data = emotion_result[0]
+        return f"Emotion: {emotion_data['label']}\nConfidence: {emotion_data['score']:.2f}"
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+def analyze_voice_emotion(audio):
+    try:
+        # Convert audio to numpy array
+        y = audio[1]
+        sr = audio[0]
+        
+        # Extract features
+        pitch = float(librosa.feature.spectral_centroid(y=y, sr=sr).mean())
+        intensity = float(librosa.feature.rms(y=y).mean())
+        tempo, _ = librosa.beat.beat_track(y=y, sr=sr)
+
+        # Simple emotion classification
+        if pitch < 150 and intensity < 0.02:
+            emotion = "sadness"
+        elif pitch > 200 and intensity > 0.05:
+            emotion = "anger"
+        elif pitch > 150 and intensity < 0.03:
+            emotion = "joy"
+        else:
+            emotion = "anxiety"
+
+        return f"Emotion: {emotion}\nPitch: {pitch:.2f}\nIntensity: {intensity:.2f}\nTempo: {tempo:.2f}"
+    except Exception as e:
+        return f"Error: {str(e)}"
+
+def chat_and_tts(message):
+    try:
+        # Get ChatGPT response
+        openai.api_key = OPENAI_API_KEY
+        chat_response = openai.ChatCompletion.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "system", "content": "You are a helpful assistant."},
+                {"role": "user", "content": message},
+            ]
+        )
+        response_text = chat_response['choices'][0]['message']['content'].strip()
+
+        # Convert to speech using Eleven Labs
+        url = f"https://api.elevenlabs.io/v1/text-to-speech/{VOICE_ID}"
+        headers = {
+            "xi-api-key": ELEVEN_LABS_API_KEY,
+            "Content-Type": "application/json"
+        }
+        data = {
+            "text": response_text,
+            "voice_settings": {
+                "stability": 0.75,
+                "similarity_boost": 0.75
+            }
+        }
+        response = requests.post(url, json=data, headers=headers)
+        
+        if response.status_code != 200:
+            return response_text, None
+            
+        # Save audio temporarily
+        audio_path = "response.mp3"
+        with open(audio_path, "wb") as f:
+            f.write(response.content)
+            
+        return response_text, audio_path
+    except Exception as e:
+        return f"Error: {str(e)}", None
+
+# Create Gradio interface
+with gr.Blocks(title="AI Therapist") as demo:
+    gr.Markdown("# AI Virtual Therapist")
+    
+    with gr.Tab("Text Emotion Analysis"):
+        text_input = gr.Textbox(label="Enter text")
+        text_button = gr.Button("Analyze Text Emotion")
+        text_output = gr.Textbox(label="Emotion Analysis Result")
+        text_button.click(analyze_text_emotion, inputs=text_input, outputs=text_output)
+    
+    with gr.Tab("Voice Emotion Analysis"):
+        audio_input = gr.Audio(label="Upload Audio", type="numpy")
+        audio_button = gr.Button("Analyze Voice Emotion")
+        audio_output = gr.Textbox(label="Voice Analysis Result")
+        audio_button.click(analyze_voice_emotion, inputs=audio_input, outputs=audio_output)
+    
+    with gr.Tab("Chat with TTS"):
+        chat_input = gr.Textbox(label="Enter your message")
+        chat_button = gr.Button("Send Message")
+        chat_output = gr.Textbox(label="Assistant Response")
+        audio_output = gr.Audio(label="Voice Response")
+        chat_button.click(chat_and_tts, inputs=chat_input, outputs=[chat_output, audio_output])
+
+# Mount Gradio app to FastAPI
+app = gr.mount_gradio_app(app, demo, path="/")

requirements.txt

@@ -0,0 +1,11 @@
+fastapi
+gradio
+uvicorn
+python-multipart
+openai
+librosa
+transformers
+torch
+requests
+python-dotenv
+soundfile