# Affinity and Diversity: Quantifying Mechanisms of Data Augmentation

arXiv: https://arxiv.org/abs/2002.08973 · HTML: https://mlsanity.com/paper/2002.08973/

## Key facts

- Cited by 45 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Feb 2020.

## Built on by (most cited first, 30 of 45)

- [A Survey of Data Augmentation Approaches for NLP](https://mlsanity.com/paper/2105.03075/) (2021)
- [SCARF: Self-Supervised Contrastive Learning using Random Feature Corruption](https://mlsanity.com/paper/2106.15147/) (2021)
- [How Much Data Are Augmentations Worth? An Investigation into Scaling Laws, Invariance, and Implicit Regularization](https://mlsanity.com/paper/2210.06441/) (2022)
- [A Comprehensive Survey on Data Augmentation](https://mlsanity.com/paper/2405.09591/) (2024)
- [MixMo: Mixing Multiple Inputs for Multiple Outputs via Deep Subnetworks](https://mlsanity.com/paper/2103.06132/) (2021)
- [Augmentations in Graph Contrastive Learning: Current Methodological Flaws & Towards Better Practices](https://mlsanity.com/paper/2111.03220/) (2021)
- [The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective](https://mlsanity.com/paper/2210.05021/) (2022)
- [Out-of-Domain Robustness via Targeted Augmentations](https://mlsanity.com/paper/2302.11861/) (2023)
- [A Theory of PAC Learnability under Transformation Invariances](https://mlsanity.com/paper/2202.07552/) (2022)
- [Barely-Supervised Learning: Semi-Supervised Learning with very few labeled images](https://mlsanity.com/paper/2112.12004/) (2021)
- [DivAug: Plug-in Automated Data Augmentation with Explicit Diversity Maximization](https://mlsanity.com/paper/2103.14545/) (2021)
- [CutDepth:Edge-aware Data Augmentation in Depth Estimation](https://mlsanity.com/paper/2107.07684/) (2021)
- [IPMix: Label-Preserving Data Augmentation Method for Training Robust Classifiers](https://mlsanity.com/paper/2310.04780/) (2023)
- [Understanding the Detrimental Class-level Effects of Data Augmentation](https://mlsanity.com/paper/2401.01764/) (2023)
- [A Survey of Automated Data Augmentation Algorithms for Deep Learning-based Image Classification Tasks](https://mlsanity.com/paper/2206.06544/) (2022)
- [Mask-then-Fill: A Flexible and Effective Data Augmentation Framework for Event Extraction](https://mlsanity.com/paper/2301.02427/) (2023)
- [Chasing Progress, Not Perfection: Revisiting Strategies for End-to-End LLM Plan Generation](https://mlsanity.com/paper/2412.10675/) (2024)
- [A Survey of Data Synthesis Approaches](https://mlsanity.com/paper/2407.03672/) (2024)
- [Local Patch AutoAugment with Multi-Agent Collaboration](https://mlsanity.com/paper/2103.11099/) (2021)
- [The Imaginative Generative Adversarial Network: Automatic Data Augmentation for Dynamic Skeleton-Based Hand Gesture and Human Action Recognition](https://mlsanity.com/paper/2105.13061/) (2021)
- [Boosting Model Inversion Attacks with Adversarial Examples](https://mlsanity.com/paper/2306.13965/) (2023)
- [Diversity-oriented Data Augmentation with Large Language Models](https://mlsanity.com/paper/2502.11671/) (2025)
- [How to Fine-tune Models with Few Samples: Update, Data Augmentation, and Test-time Augmentation](https://mlsanity.com/paper/2205.07874/) (2022)
- [Semi-Supervised Learning for Multi-Label Cardiovascular Diseases Prediction:A Multi-Dataset Study](https://mlsanity.com/paper/2306.10494/) (2023)
- [Principled Ultrasound Data Augmentation for Classification of Standard Planes](https://mlsanity.com/paper/2103.07895/) (2021)
- [Domain Generalization by Rejecting Extreme Augmentations](https://mlsanity.com/paper/2310.06670/) (2023)
- [Local Superior Soups: A Catalyst for Model Merging in Cross-Silo Federated Learning](https://mlsanity.com/paper/2410.23660/) (2024)
- [Measuring Diversity in Synthetic Datasets](https://mlsanity.com/paper/2502.08512/) (2025)
- [On Data-centric Myths](https://mlsanity.com/paper/2111.11514/) (2021)
- [Data Diversity as Implicit Regularization: How Does Diversity Shape the Weight Space of Deep Neural Networks?](https://mlsanity.com/paper/2410.14602/) (2024)

## Abstract (arXiv)

Though data augmentation has become a standard component of deep neural network training, the underlying mechanism behind the effectiveness of these techniques remains poorly understood. In practice, augmentation policies are often chosen using heuristics of either distribution shift or augmentation diversity. Inspired by these, we seek to quantify how data augmentation improves model generalization. To this end, we introduce interpretable and easy-to-compute measures: Affinity and Diversity. We find that augmentation performance is predicted not by either of these alone but by jointly optimizing the two.