This tool visualizes bounding box coordinates returned by the Gemini API for uploaded images. Upload an image, enter a prompt asking for object detection, and the application will process your request using the selected Gemini model. The tool displays the original image with colored bounding boxes overlaid on a coordinate grid (0-1000), and extracts each detected region as a separate image for detailed inspection.