Image & Computer Vision datasets

Image datasets for object detection, segmentation, document understanding and visual recognition. Listings include street scenes, retail imagery, documents and specialised domains, with bounding boxes, polygons or labels where provided.

The catalogue lists 591 image & computer vision datasets. Most listings are image, video and medical data, typically used for document understanding & OCR, object detection and LLM pre-training. Where providers state a volume in hours, it ranges from 1,000 hours to 100k hours.