Computer Vision
1.SimCLR: A Simple Framework for Contrastive Learning of Visual Representations
Abstract & Key Attribution contrast learning에서, (1) composition of data augmentation, (2) learnable non-linear transformation between representation
2024년 11월 22일