Self-Supervised Learning: A method of unsupervised learning, where the model trains on the input data without any explicit labels, it’s mainly used for representation learning and transfer learning.