This morning, Unsplash announced that it was making its dataset completely open source. With over 200,000 contributing photographers across the globe, Unsplash has access to over two million images.
With this announcement, Unsplash is releasing the dataset for people and companies to use for research in machine learning, image quality, search engines and more. All of the images will be of high quality and contain attributes like EXIF data, image categories and subcategories, keywords, stats on viewing and more. The images will also list the original contributor.
Two versions of the dataset will be available — a “lite” version for commercial and noncommercial use, with 25,000 images available for download. A full version will also be available for noncommercial use, with the full library of over two million images.