# Self-Augmentation: Generalizing Deep Networks to Unseen Classes for Few-Shot Learning

arXiv: https://arxiv.org/abs/2004.00251 · HTML: https://mlsanity.com/paper/2004.00251/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Apr 2020.

## Built on by (most cited first, 9 of 9)

- [Image Data Augmentation Approaches: A Comprehensive Survey and Future directions](https://mlsanity.com/paper/2301.02830/) (2023)
- [Local Propagation for Few-Shot Learning](https://mlsanity.com/paper/2101.01480/) (2021)
- [Ortho-Shot: Low Displacement Rank Regularization with Data Augmentation for Few-Shot Learning](https://mlsanity.com/paper/2110.09374/) (2021)
- [KeepOriginalAugment: Single Image-based Better Information-Preserving Data Augmentation Approach](https://mlsanity.com/paper/2405.06354/) (2024)
- [Few-Shot Object Detection with Proposal Balance Refinement](https://mlsanity.com/paper/2204.10527/) (2022)
- [Local-Global Temporal Fusion Network with an Attention Mechanism for Multiple and Multiclass Arrhythmia Classification](https://mlsanity.com/paper/2308.02416/) (2023)
- [ProCal: Inference-Time Proposal Calibration for Open-Vocabulary Object Detection](https://mlsanity.com/paper/2607.01759/) (2026)
- [DiCE-CIR: Direct Composition Learning for Efficient Zero-Shot Composed Image Retrieval](https://mlsanity.com/paper/2607.04665/) (2026)
- [Tree-of-Thoughts Reasoning for Text-to-Image In-Context Learning](https://mlsanity.com/paper/2607.07117/) (2026)

## Abstract (arXiv)

Few-shot learning aims to classify unseen classes with a few training examples. While recent works have shown that standard mini-batch training with a carefully designed training strategy can improve generalization ability for unseen classes, well-known problems in deep networks such as memorizing training statistics have been less explored for few-shot learning. To tackle this issue, we propose self-augmentation that consolidates self-mix and self-distillation. Specifically, we exploit a regional dropout technique called self-mix, in which a patch of an image is substituted into other values in the same image. Then, we employ a backbone network that has auxiliary branches with its own classifier to enforce knowledge sharing. Lastly, we present a local representation learner to further exploit a few training examples for unseen classes. Experimental results show that the proposed method outperforms the state-of-the-art methods for prevalent few-shot benchmarks and improves the generalization ability.