4M-21: An Any-to-Any Vision Model for Tens of Tasks and Modalities – Apple Machine Learning Research
4M-21: An Any-to-Any Vision Model for Tens of Tasks and Modalities – Apple Machine Learning Research