Interpolation Consistency Training for Semi-Supervised Learning

Vikas Verma, Kenji Kawaguchi, Alex Lamb, Juho Kannala, Arno Solin +2 more
3/9/2019
stat.MLcs.AIcs.LG

Official code: vikasverma1077/ICT (151 stars, last commit 2022-10-16).

Abstract

We introduce Interpolation Consistency Training (ICT), a simple and computation efficient algorithm for training Deep Neural Networks in the semi-supervised learning paradigm. ICT encourages the prediction at an interpolation of unlabeled points to be consistent with the interpolation of the predictions at those points. In classification problems, ICT moves the decision boundary to low-density regions of the data distribution. Our experiments show that ICT achieves state-of-the-art performance when applied to standard neural network architectures on the CIFAR-10 and SVHN benchmark datasets. Our theoretical analysis shows that ICT corresponds to a certain type of data-adaptive regularization with unlabeled points which reduces overfitting to labeled points under high confidence values.

View on arXivView PDF

Code Implementations(1)

vikasverma1077/ICTOfficial100%

Code for reproducing ICT (published in Neural Networks 2022, and in IJCAI 2019)

15133PythonMar 5, 20194 years ago
deep-learningdeep-neural-networkssemi-supervised-learning

Similar Papers

Temporal Ensembling for Semi-Supervised Learning

86% match1 repo1612016

Exploring Beyond Logits: Hierarchical Dynamic Labeling Based on Embeddings for Semi-Supervised Classification

86% match2024

Semi-Supervised Learning via Swapped Prediction for Communication Signal Recognition

85% match2023

NP-Match: When Neural Processes meet Semi-Supervised Learning

85% match1 repo962022

Boosting Semi-Supervised Learning by bridging high and low-confidence predictions

85% match2023

An Uncertainty-Aware Pseudo-Label Selection Framework using Regularized Conformal Prediction

85% match1 repo72023

Cite this paper

@article{verma2019interpolation,
  title  = {Interpolation Consistency Training for Semi-Supervised Learning},
  author = {Vikas Verma and Kenji Kawaguchi and Alex Lamb and Juho Kannala and Arno Solin and Yoshua Bengio and David Lopez-Paz},
  year   = {2019},
  eprint = {1903.03825},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1903.03825}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1903.03825.svg)](https://codeofpaper.com/paper/1903.03825)

Discussion