app.py

import gradio as gr
from keras.models import load_model
from PIL import Image, ImageOps
import numpy as np
from huggingface_hub import InferenceClient

# Load the Keras model
model = load_model("keras_model.h5", compile=False)

# Load class labels from a file
with open("labels.txt", "r") as file:
    class_names = [line.strip() for line in file]

# Initialize the Hugging Face Inference Client
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")

def classify_image(img):
    # Resize and normalize the image for model prediction
    image = ImageOps.fit(img, (224, 224), Image.Resampling.LANCZOS)
    image_array = np.asarray(image)
    normalized_image_array = (image_array.astype(np.float32) / 127.5) - 1
    data = normalized_image_array.reshape((1, 224, 224, 3))

    # Predict the emotion using the model
    prediction = model.predict(data)
    index = np.argmax(prediction)
    class_name = class_names[index]
    confidence_score = prediction[0][index]

    # Return the detected emotion and confidence score
    return {
        "Detected Emotion": class_name,
        "Confidence Score": f"{confidence_score:.2f}"
    }

def respond(
    messages,
    system_message,
    max_tokens,
    temperature,
    top_p
):
    # Ensure messages are correctly formatted
    formatted_messages = []
    for message in messages:
        if message['content'] is None:
            message['content'] = ''  # Set to empty string if None
        formatted_messages.append(message)

    # Add system message at the beginning
    formatted_messages.insert(0, {"role": "system", "content": system_message})

    # Proceed with chat completion
    try:
        response = ""
        for message in client.chat_completion(
            model_id='HuggingFaceH4/zephyr-7b-beta',
            messages=formatted_messages,
            max_tokens=max_tokens,
            temperature=temperature,
            top_p=top_p,
            stream=True
        ):
            token = message.choices[0].delta.content
            response += token
        return response
    except Exception as e:
        print(f"Error: {e}")
        return "Sorry, there was an error processing your request."

def emotion_and_chat(img, system_message, max_tokens, temperature, top_p):
    # Classify the image to detect emotion
    emotion_result = classify_image(img)
    detected_emotion = emotion_result["Detected Emotion"]

    # Start chatbot conversation based on the detected emotion
    initial_message = f"I detected that you're feeling {detected_emotion}. Let's talk about it."
    chat_history = [{"role": "user", "content": initial_message}]
    chat_response = respond(chat_history, system_message, max_tokens, temperature, top_p)

    return chat_response

# Define custom CSS for styling
custom_css = """
body {
    font-family: Arial, sans-serif;
    background-color: #000;
    color: #f4f4f4;
}
.gradio-container {
    border-radius: 10px;
    padding: 20px;
    background: linear-gradient(135deg, #ff0000, #008000);
    box-shadow: 0 4px 15px rgba(0, 0, 0, 0.2);
}
.gradio-container h1 {
    font-family: Arial, sans-serif;
    font-size: 2.5em;
    text-align: center;
    color: #fff;
}
.gradio-container p {
    font-size: 1em;
    text-align: center;
    color: #c0c0c0;
}
.gradio-button {
    background-color: #ff0000;
    border: none;
    color: #fff;
    padding: 10px 20px;
    font-size: 1em;
    cursor: pointer;
    border-radius: 5px;
    transition: background-color 0.2s ease;
}
.gradio-button:hover {
    background-color: #ff4d4d;
}
#output-container {
    border-radius: 10px;
    background-color: #008000;
    padding: 20px;
    color: #fff;
}
#output-container h3 {
    font-family: Arial, sans-serif;
    font-size: 1.5em;
    color: #fff;
}
.gr-examples {
    text-align: center;
}
.gr-example-img {
    width: 100px;
    border-radius: 5px;
    margin: 5px;
    box-shadow: 0 4px 10px rgba(0, 0, 0, 0.2);
}
"""

# Define example images from URLs
examples = [
    "https://firebasestorage.googleapis.com/v0/b/hisia-4b65b.appspot.com/o/a-captivating-ukiyo-e-inspired-poster-featuring-a--wTg7L-f2Tfiy6K8w6aWnKA-KbGU9GSKSDGBbbxrCO65Mg.jpeg?alt=media&token=64590de9-e265-44ac-a766-aeecd455ed5d",
    "https://firebasestorage.googleapis.com/v0/b/hisia-4b65b.appspot.com/o/poster-ai-themed-kenyan-female-silhoutte-written-l-PMIXpNWGQ8KaNNetQRVJuQ-B1TteyL-S5OTPZFXvfGybg.jpeg?alt=media&token=fc10f96d-403e-4f75-bd9c-810e0da36867",
    "https://firebasestorage.googleapis.com/v0/b/hisia-4b65b.appspot.com/o/poster-ai-themed-kenyan-male-silhoutte-written-log-z3fqBD5bQOOj6uqGd_iXLQ-4aBfNy0ZTgmLlTsZh1dzIA.jpeg?alt=media&token=f218f160-d38e-482f-97a9-5442c2f251a7"
]

# Gradio Interface
interface = gr.Interface(
    fn=emotion_and_chat,
    inputs=[
        gr.Image(type="pil", label="Upload an Image"),
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
    ],
    outputs=gr.Chatbot(label="Chat with the AI"),
    examples=examples,
    title="HISIA: Emotion Detector and Chatbot",
    description="Upload an image, and our AI will detect the emotion expressed in it and start a conversation with you.",
    allow_flagging="never",
    css=custom_css,
)

# Launch the Gradio interface
if __name__ == "__main__":
    interface.launch()