Analyze images to generate captions, detect objects, or perform OCR

Enhance and upscale images with face restoration

Segment objects in images and videos using text prompts

Generate depth maps from images

Tag images with labels

FitDiT is a high-fidelity virtual try-on model.

Convert images of screens to structured elements

Generate clickable coordinates on a screenshot

Interact with Florence-2 to analyze images and generate descriptions

Find similar images from a collection

Generate 3D depth maps from images and videos

Generate depth map from images

Extract text from images using OCR

Find similar images by uploading a photo

Segment body parts in images

Art Institute of Chicago Gallery

Meta Llama3 8b with Llava Multimodal capabilities

Find images matching a text query

Compute normals for images and videos

Vote on background-removed images to rank models