app.py

import gradio as gr
from typing import Any, Dict, List

from src.registry import get_model_display_names, get_model

APP_TITLE = "Machine Learning CS 6140 Project: Pet Recognizer"
TOP_K_DEFAULT = 5

DARK_CSS = """
body {
    background-color: #0f172a !important;
}

.gradio-container {
    background-color: #0f172a !important;
    color: #e5e7eb !important;
}

h1, h2, h3, h4, p, li, label {
    color: #e5e7eb !important;
}

a {
    color: #60a5fa !important;
}

.gr-box {
    background-color: #020617 !important;
    border-radius: 10px;
}

.gr-button {
    background-color: #1e293b !important;
    color: #e5e7eb !important;
}

.gr-button:hover {
    background-color: #334155 !important;
}
"""


# -----------------------------
# Helpers
# -----------------------------
def format_topk_for_table(top_k: List[Dict[str, Any]]) -> List[List[Any]]:
    rows = []
    for rank, entry in enumerate(top_k, start=1):
        class_name = entry.get("class_name", f"id={entry.get('class_id', '?')}")
        prob = entry.get("probability", 0.0)
        rows.append([rank, class_name, round(float(prob) * 100.0, 2)])
    return rows


def run_inference(model_id: str, image) -> Dict[str, Any]:
    if image is None:
        return {
            "main_text": "Please upload an image first.",
            "topk_table": [],
        }

    model = get_model(model_id)
    result = model.predict(image, top_k=TOP_K_DEFAULT)

    class_name = result.get("class_name", "Unknown")
    class_id = result.get("class_id", "N/A")
    top_k = result.get("top_k", [])

    main_text = (
        f"**Predicted Class:** {class_name}  \n"
        f"**Class ID:** {class_id}"
    )

    return {
        "main_text": main_text,
        "topk_table": format_topk_for_table(top_k),
    }


# -----------------------------
# UI
# -----------------------------
def build_demo() -> gr.Blocks:
    model_display_names = get_model_display_names()
    name_to_id = {v: k for k, v in model_display_names.items()}
    default_display_name = next(iter(name_to_id.keys()))

    with gr.Blocks(css=DARK_CSS) as demo:

        # Title
        gr.Markdown(
            f"""
#  {APP_TITLE}

This project demonstrates **pet breed recognition** using the  
**Oxford-IIIT Pet Dataset**, comparing **classical machine learning models**
(Logistic Regression, SVM) with **deep feature-based models**
(Pretrained ResNet18).

 **Dataset & Supported Breeds**  
The models are trained on **37 cat and dog breeds** from the Oxford-IIIT Pet Dataset.  
 https://www.robots.ox.ac.uk/~vgg/data/pets/
"""
        )

        # Instructions
        gr.Markdown(
            """
##  Instructions

1. **Upload** a clear, close-up image of a **cat or dog** belonging to one of the supported breeds  
2. **Select a model** to run the recognition:
   - **LR / SVM** → Expected to perform poorly on raw pixel inputs  
   - **ResNet-based models** → Use pretrained deep visual features and produce much better results
3. Click **Run Identification** to view the **Top-5 predictions**
"""
        )

        with gr.Row():
            # Left column
            with gr.Column(scale=1):
                gr.Markdown("### Select Model & Upload Image")

                model_dropdown = gr.Dropdown(
                    choices=list(name_to_id.keys()),
                    value=default_display_name,
                    label="Select Model",
                )

                image_input = gr.Image(
                    type="pil",
                    label="Upload your pet image (JPEG / PNG)",
                )

                run_button = gr.Button("Run Identification")

            # Right column
            with gr.Column(scale=1):
                gr.Markdown("### Model Prediction")

                main_output = gr.Markdown(
                    value="Prediction will appear here.",
                )

                topk_output = gr.Dataframe(
                    headers=["Rank", "Class Name", "Probability (%)"],
                    datatype=["number", "str", "number"],
                    column_count=3,
                    label=f"Top-{TOP_K_DEFAULT} Predictions",
                )

        # Button wiring
        def _gradio_infer(selected_display_name, img):
            model_id = name_to_id[selected_display_name]
            result = run_inference(model_id, img)
            return result["main_text"], result["topk_table"]

        run_button.click(
            fn=_gradio_infer,
            inputs=[model_dropdown, image_input],
            outputs=[main_output, topk_output],
        )

    return demo


if __name__ == "__main__":
    demo = build_demo()
    demo.launch()