Last week, Getty Images, the world's largest commercial image library, announced a significant initiative: providing AI developers with a high-quality photo dataset for training purposes. This dataset includes 3,750 high-quality photos across 15 categories such as business, education, healthcare, sports and fitness, objects, illustrations, and icons, aiming to support the development and training of various AI applications.
The dataset is now available on the Hugging Face platform, allowing users to access it for free after agreeing to the terms of service and providing their contact information. Getty Images' move is intended to attract businesses and developers with free samples while promoting its paid licensing services.
The company claims to have over 572 million photos, with more than 200 million available for commercial use. Each photo comes with structured metadata, including information such as age and gender, with an average of 50 keywords per image. This not only aids in the training of AI models but also provides users with secure training resources, reducing the risk of infringement.
This initiative reflects a proactive transformation of traditional image library services towards the AI era, injecting new momentum into the development of image recognition and generation technologies. Getty Images' strategic move could have a profound impact on the industry, promoting innovation in the integration of AI with traditional image library services.
As AI technology rapidly advances in image processing and generation, high-quality training data will become a valuable resource for developers. Getty Images' move undoubtedly sets a new benchmark for the industry.
Address: https://github.com/mayuelala/FollowYourCanvas