US Patent Application 18343579. Systems and Methods for Contrastive Learning of Visual Representations simplified abstract
Contents
Systems and Methods for Contrastive Learning of Visual Representations
Organization Name
Inventor(s)
Simon Komblith of Toronto (CA)
Mohammad Norouzi of Toronto (CA)
Geoffrey Everest Hinton of Toronto (CA)
Kevin Jordan Swersky of Mississauga (CA)
Systems and Methods for Contrastive Learning of Visual Representations - A simplified explanation of the abstract
- This abstract for appeared for US patent application number 18343579 Titled 'Systems and Methods for Contrastive Learning of Visual Representations'
Simplified Explanation
This abstract describes a method for improving visual representations using semi-supervised contrastive learning. The method involves using specific data augmentation techniques and a learnable transformation to enhance the visual representations. It also includes improvements for semi-supervised contrastive learning, such as generating an image classification model based on unlabeled training data, fine-tuning the model using labeled training data, and distilling the model into a smaller student model with fewer parameters.
Original Abstract Submitted
Systems, methods, and computer program products for performing semi-supervised contrastive learning of visual representations are provided. For example, the present disclosure provides systems and methods that leverage particular data augmentation schemes and a learnable nonlinear transformation between the representation and the contrastive loss to provide improved visual representations. Further, the present disclosure also provides improvements for semi-supervised contrastive learning. For example, computer-implemented method may include performing semi-supervised contrastive learning based on a set of one or more unlabeled training data, generating an image classification model based on a portion of a plurality of layers in a projection head neural network used in performing the contrastive learning, performing fine-tuning of the image classification model based on a set of one or more labeled training data, and after performing the fine-tuning, distilling the image classification model to a student model comprising a relatively smaller number of parameters than the image classification model.
