app.py

import os
import time
import gradio as gr
from typing import *
from pillow_heif import register_heif_opener
register_heif_opener()
import vision_agent as va
from vision_agent.tools import register_tool
from vision_agent.tools import load_image, owl_v2, overlay_bounding_boxes, save_image

from huggingface_hub import login
import spaces

# Perform login using the token
hf_token = os.getenv("HF_TOKEN")
login(token=hf_token, add_to_git_credential=True)

def detect_brain_tumor(image, debug: bool = False) -> str:
    """
    Detects a brain tumor in the given image and saves the image with bounding boxes.

    Parameters:
        image: The input image (as provided by Gradio).
        debug (bool): Flag to enable logging for debugging purposes.

    Returns:
        str: Path to the saved output image.
    """
    # Generate a unique output filename
    output_path = f"./output/tumor_detection_{int(time.time())}.jpg"
    
    if debug:
        print(f"Image received")

    # Step 2: Detect brain tumor using owl_v2
    prompt = "detect brain tumor"
    detections = owl_v2(prompt, image)
    if debug:
        print(f"Detections: {detections}")

    # Step 3: Overlay bounding boxes on the image
    image_with_bboxes = overlay_bounding_boxes(image, detections)
    if debug:
        print("Bounding boxes overlaid on the image")

    # Step 4: Save the resulting image
    save_image(image_with_bboxes, output_path)
    if debug:
        print(f"Image saved to {output_path}")
    
    return output_path

INTRO_TEXT="# 🔬🧠 CellVision AI -- Intelligent Cell Imaging Analysis 🤖🧫"
IMAGE_PROMPT="Are these cells healthy or cancerous?"

with gr.Blocks(css="style.css") as demo:
    gr.Markdown(INTRO_TEXT)
    with gr.Tab("Agentic Detection"):
        with gr.Row():
            with gr.Column():
                image = gr.Image(type="numpy")
            with gr.Column():
                text_input = gr.Text(label="Input Text")
                text_output = gr.Text(label="Text Output")
        chat_btn = gr.Button()

        chat_inputs = [
            image
        ]
        chat_outputs = [
            text_output
        ]
        chat_btn.click(
            fn=detect_brain_tumor,
            inputs=chat_inputs,
            outputs=chat_outputs,
        )
        
        examples = [["./examples/194_jpg.rf.3e3dd592d034bb5ee27a978553819f42.jpg"],
                    ["./examples/239_jpg.rf.3dcc0799277fb78a2ab21db7761ccaeb.jpg"],
                    ["./examples/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg"],
                    ["./examples/1491_jpg.rf.3c658e83538de0fa5a3f4e13d7d85f12.jpg"],
                    ["./examples/1550_jpg.rf.3d067be9580ec32dbee5a89c675d8459.jpg"],
                    ["./examples/2256_jpg.rf.3afd7903eaf3f3c5aa8da4bbb928bc19.jpg"],
                    ["./examples/2871_jpg.rf.3b6eadfbb369abc2b3bcb52b406b74f2.jpg"],
                    ["./examples/2921_jpg.rf.3b952f91f27a6248091e7601c22323ad.jpg"],
                    ]
        gr.Examples(
            examples=examples,
            inputs=chat_inputs,
        )

if __name__ == "__main__":
    demo.queue(max_size=10).launch(debug=True)