Research map / Label-efficient and robust learning / Recognition and long-tail
Data Augmentation as Feature Manipulation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Balance the feature views through data augmentation | Two-layer patch-wise CNN trained by gradient descent · Multi-view data model | test misclassification probability: \geq\left(\frac{1}{2}-O\left(\frac 1 {\sqrt{C}}\right)\right)\sum_{k=2}^{K}\rho_{k} -> \leq\frac{nK}{\textup{poly} (d)} | Helped |
| Randomly shift the spurious feature position during augmentation | CIFAR-10 cat-versus-dog classification · CIFAR-10 | test accuracy | Helped |
| Use random crop rather than center crop | CIFAR-10 cat-versus-dog classification · CIFAR-10 | test accuracy | Helped |
| Replace independent training samples with fixed augmented variants | ResNet20 · CIFAR-10 test set | test accuracy | Helped |
| Add extra samples from the dominant view to a balanced dataset | Convolutional network on a synthetic unbalanced-view dataset · Synthetic unbalanced-view dataset | test accuracy | Mixed |
Closest papers by meaning
- A Kernel Theory of Modern Data Augmentation (ICML 2019 · similarity 0.90)
- Data-Efficient Augmentation for Training Neural Networks (NeurIPS 2022 · similarity 0.89)
- Automatic Data Augmentation via Invariance-Constrained Learning (ICML 2023 · similarity 0.89)
- Tradeoffs in Data Augmentation: An Empirical Study (ICLR 2021 · similarity 0.88)
- How Data Augmentation Shapes Neural Representations (NeurIPS 2026 · similarity 0.86)
- Characterizing the Generalization Error of Random Feature Regression with Arbitrary Data-Augmentation (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.