Papers of the day   All papers

FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

Comments

Jan 22 2020 Ian Goodfellow

The quiet semisupervised revolution continues
7 replies, 1102 likes


Jan 22 2020 David Berthelot

FixMatch: focusing on simplicity for semi-supervised learning and improving state of the art (CIFAR 94.9% with 250 labels, 88.6% with 40). https://arxiv.org/abs/2001.07685 Collaboration with Kihyuk Sohn, @chunliang_tw @ZizhaoZhang Nicholas Carlini @ekindogus @Han_Zhang_ @colinraffel https://t.co/BmeYvpEHzX
5 replies, 937 likes


Feb 05 2020 Alexey Kurakin

Fixmatch: code for training on Imagenet dataset is released and available here: https://github.com/google-research/fixmatch/tree/master/imagenet https://arxiv.org/abs/2001.07685 by Kihyuk Sohn @D_Berthelot_ML @chunliang_tw @ZizhaoZhang Nicholas Carlini @ekindogus @alexey2004 @Han_Zhang_ @colinraffel
0 replies, 261 likes


Jan 23 2020 hardmaru

Happy to see SOTA benchmarks moving from CIFAR-10, to CIFAR-10 with only 40 training labels! https://arxiv.org/abs/2001.07685
1 replies, 139 likes


Jan 01 2020 Connor Shorten

This is my recap of Artificial Intelligence in 2019! This video covers new developments in understanding Neural Networks, self-supervised learning, language models, Generative Models, Game-playing RL, and many more! https://youtu.be/6SWpN64Ivb4 #100DaysOfMLCode
0 replies, 128 likes


Feb 05 2020 Yann N. Dauphin

Ten years ago, the mcRBM really crushed it with an amazing 71% accuracy on CIFAR-10. Today, Fixmatch reaches 88.6% with 1000x fewer labelled examples. https://arxiv.org/pdf/2001.07685.pdf @D_Berthelot_ML #ML10YearChallenge https://t.co/dKr3Qup76Y
0 replies, 95 likes


Jan 22 2020 Aakash Kumar Nain

Paper for the day!
0 replies, 16 likes


Jan 24 2020 mat kelcey

Awesome example of how to stitch together a semi supervised learning pipeline! Makes me wonder how these operators; pseudo label, dataset union, augment, etc; could be represented and "learnt" through evolutionary approaches or framed as RL... @D_Berthelot_ML doable?
1 replies, 11 likes


Jan 23 2020 Aakash Kumar Nain

Why I liked this paper so much? The "ablation study" section is so clearly written! Thanks @D_Berthelot_ML @colinraffel et al
1 replies, 10 likes


Jan 22 2020 phalanx

FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence paper: https://arxiv.org/abs/2001.07685 code: https://github.com/google-research/fixmatch https://t.co/9htNCgAnlN
0 replies, 8 likes


Jan 22 2020 Daisuke Okanohara

FixMatch achieves new SOTA performance on semi-supervised learning tasks. 1) generate pseudo-labels with weak-augmentation and keep high confidence samples 2) train the model using generated training data with strong augmentation. https://arxiv.org/abs/2001.07685
0 replies, 8 likes


Feb 13 2020 Alexey Romanov

FixMatch https://arxiv.org/abs/2001.07685 gets even better 88.61% accuracy in the same situation #MachineLearning
0 replies, 1 likes


Jan 29 2020 彡Sαι彡

Thanks @CShorten30 @labs_henry My top -5 are, 1. Fixmatch - SSL @GoogleAI https://arxiv.org/pdf/2001.07685.pdf 2. Scaling laws of Neural language models @OpenAI https://arxiv.org/abs/2001.08361 3. Squinting at VQA models @MSFTResearch https://arxiv.org/abs/2001.06927 4. Jax&Swift http://inoryy.com/post/next-gen-ml-tools/
1 replies, 0 likes


Content