mlsanityData to 7 Oct 2026

AdMix: A Mixed Sample Data Augmentation Method for Neural Machine Translation

arXiv 2205.04686 ↗First posted 10 May 2022cs.CLcited by 4

Not on the map. Why?

Abstract (arXiv)

In Neural Machine Translation (NMT), data augmentation methods such as back-translation have proven their effectiveness in improving translation performance. In this paper, we propose a novel data augmentation approach for NMT, which is independent of any additional training data. Our approach, AdMix, consists of two parts: 1) introduce faint discrete noise (word replacement, word dropping, word swapping) into the original sentence pairs to form augmented samples; 2) generate new synthetic training data by softly mixing the augmented samples with their original samples in training corpus. Experiments on three translation datasets of different scales show that AdMix achieves signifi cant improvements (1.0 to 2.7 BLEU points) over strong Transformer baseline. When combined with other data augmentation techniques (e.g., back-translation), our approach can obtain further improvements.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 23 of its 29 references that are in the map. First posted on arXiv 10 May 2022.

Builds on

23 of 29 references.

References · 23

  1. 2017Attention Is All You Needcited by 59,932
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  3. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
  4. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,623
  5. 2019Unsupervised Data Augmentation for Consistency Trainingcited by 1,057
  6. 2019EDA: Easy Data Augmentation Techniques for Boosting Performance on Text Classification Taskscited by 800
  7. 2015Improving Neural Machine Translation Models with Monolingual Datacited by 784
  8. 2016Adversarial Training Methods for Semi-Supervised Text Classificationcited by 280
  9. 2017Unsupervised Machine Translation Using Monolingual Corpora Onlycited by 272
  10. 2018Adversarial Logit Pairingcited by 224
  11. 2019Not Enough Data? Deep Learning to the Rescue!cited by 174
  12. 2020Data Augmentation using Pre-trained Transformer Modelscited by 168
11 more
  1. 2019Revisiting Self-Training for Neural Sequence Generationcited by 154
  2. 2017Unsupervised Neural Machine Translationcited by 148
  3. 2021Data Augmentation Approaches in Natural Language Processing: A Surveycited by 82
  4. 2017Data Noising as Smoothing in Neural Network Language Modelscited by 78
  5. 2018Semi-Supervised Sequence Modeling with Cross-View Trainingcited by 75
  6. 2018SwitchOut: an Efficient Data Augmentation Algorithm for Neural Machine Translationcited by 72
  7. 2020Sequence-Level Mixed Sample Data Augmentationcited by 71
  8. 2020AdvAug: Robust Adversarial Augmentation for Neural Machine Translationcited by 66
  9. 2017Classical Structured Prediction Losses for Sequence to Sequence Learningcited by 55
  10. 2019Soft Contextual Data Augmentation for Neural Machine Translationcited by 49
  11. 2018Towards Robust Neural Machine Translationcited by 35

Built on by

Cited by year

Citing papers per year: 2023: 1; 2024: 2; 2025: 1.

4 papers.

  1. 2024A Survey on Mixup Augmentations and Beyondcited by 18
  2. 2023GDA: Generative Data Augmentation Techniques for Relation Extraction Taskscited by 10
  3. 2025Data Augmentation With Back translation for Low Resource languages: A case of English and Lugandacited by 3
  4. 2024Non-Fluent Synthetic Target-Language Data Improve Neural Machine Translationcited by 0