Find similar images using tags and images
Convert images of screens to structured elements
Visualize attention maps for images using selected models
Display a heat map on an interactive map
Art Institute of Chicago Gallery
Interact with Florence-2 to analyze images and generate descriptions
Display interactive UI theme preview with Gradio
Search for images or video frames online
Detect objects in images and highlight them
Vote on background-removed images to rank models
Generate saliency maps from RGB and depth images
Detect and compare dominant colors in images
Find images matching a text query
Danbooru2022 Embeddings Playground is an experimental tool designed to explore and find similar images using tags and image embeddings. It leverages advanced embedding technology to enable users to search for images based on semantic similarity, allowing for intuitive and powerful image discovery.
What is an embedding?
An embedding is a vector representation of an image or text that captures its semantic meaning in a high-dimensional space.
Can I use the tool without an account?
Yes, no account is required to use the Danbooru2022 Embeddings Playground.
How are similar images determined?
Similar images are determined by calculating the similarity score between the embedding of your input (tag or image) and the embeddings of images in the dataset.