We present Meta Pseudo Labels, a semi-supervised learning method that achieves a new state-of-the-art top-1 accuracy of 90. 2% on ImageNet, which is 1. 6% better than the existing state-of-the-art.
During the learning of the student, we inject noise such as dropout, stochastic depth, and data augmentation via RandAugment to the student so that the student generalizes better than the teacher.
Ranked #14 on Image Classification on ImageNet ReaL (using extra training data)
In this work, we present a new perspective on how to effectively noise unlabeled examples and argue that the quality of noising, specifically those produced by advanced data augmentation methods, plays a crucial role in semi-supervised learning.
Ranked #1 on Sentiment Analysis on Amazon Review Full
The human mind is a powerful multifunctional knowledge storage and management system that performs generalization, type inference, anomaly detection, stereotyping, and other tasks.
Mixture of Softmaxes (MoS) has been shown to be effective at addressing the expressiveness limitation of Softmax-based models.
Ranked #18 on Machine Translation on WMT2014 English-French
In this work, we study the credit assignment problem in reward augmented maximum likelihood (RAML) learning, and establish a theoretical equivalence between the token-level counterpart of RAML and the entropy regularized reinforcement learning.
Learning meaningful representations that maintain the content necessary for a particular task while filtering away detrimental variations is a problem of great interest in machine learning.
Knowledge bases are important resources for a variety of natural language processing tasks but suffer from incompleteness.
Dialogue state tracking (DST) is a process to estimate the distribution of the dialogue states as a dialogue progresses.