This application visualizes object detection results by drawing bounding boxes on images using the Google's Gemini 1.5 Pro AI model.
API keys are only stored in your browser's local storage.
Images are only sent to Google's Gemini API servers for processing.
Try the following images (click the links to add them):
Sometimes getting a description of the image helps you understand what the AI sees. This can help you better set up the prompt and settings.
Use this to detect bounding boxes of objects in an image. Use the image description to help with setting up the prompt. Rows and columns determine how many tiles the image will be divided into — for simple images, 1x1 is best.
For the free API you might get the "Resource has been exhausted" error if you make too many requests too quickly.
This experiment uses a lot of ideas and code from Simon Willison's original post which you can read here: https://simonwillison.net/2024/Aug/26/gemini-bounding-box-visualization/
