simclrv2

SimCLRv2 - Big Self-Supervised Models are Strong Semi-Supervised Learners

Self-Supervised Learning in PyTorch

[CVPR 2021] "The Lottery Tickets Hypothesis for Supervised and Self-supervised Pre-training in Computer Vision Models" Tianlong Chen, Jonathan Frankle, Shiyu Chang, Sijia Liu, Yang Zhang, Michael Carbin, Zhangyang Wang

Finetuning and clustering library for image perceptual similarity models.

This is an implementation of SimCLR for Anime Contrastive Learning. The goal of this project is to build a model that understand the general feature of anime compared to sketch using semi-supervised learning, letting the model to learn how to classify which image is anime or not.