# On-manifold Adversarial Data Augmentation Improves Uncertainty Calibration

arXiv: https://arxiv.org/abs/1912.07458 · HTML: https://mlsanity.com/paper/1912.07458/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Dec 2019.

## Built on by (most cited first, 18 of 18)

- [A Survey of Uncertainty in Deep Neural Networks](https://mlsanity.com/paper/2107.03342/) (2021)
- [A Review of Uncertainty Estimation and its Application in Medical Imaging](https://mlsanity.com/paper/2302.08119/) (2023)
- [Why Does Little Robustness Help? A Further Step Towards Understanding Adversarial Transferability](https://mlsanity.com/paper/2307.07873/) (2023)
- [Adaptive Temperature Scaling for Robust Calibration of Deep Neural Networks](https://mlsanity.com/paper/2208.00461/) (2022)
- [PnPOOD : Out-Of-Distribution Detection for Text Classification via Plug andPlay Data Augmentation](https://mlsanity.com/paper/2111.00506/) (2021)
- [Understanding Adversarial Robustness Against On-manifold Adversarial Examples](https://mlsanity.com/paper/2210.00430/) (2022)
- [A Structured Review of Literature on Uncertainty in Machine Learning & Deep Learning](https://mlsanity.com/paper/2406.00332/) (2024)
- [Adversarial Purification with the Manifold Hypothesis](https://mlsanity.com/paper/2210.14404/) (2022)
- [Calibration Attacks: A Comprehensive Study of Adversarial Attacks on Model Confidence](https://mlsanity.com/paper/2401.02718/) (2024)
- [Data-Centric Digital Agriculture: A Perspective](https://mlsanity.com/paper/2312.03437/) (2023)
- [Improving Uncertainty of Deep Learning-based Object Classification on Radar Spectra using Label Smoothing](https://mlsanity.com/paper/2109.12851/) (2021)
- [Investigation of Uncertainty of Deep Learning-based Object Classification on Radar Spectra](https://mlsanity.com/paper/2106.05870/) (2021)
- [Generalized Uncertainty of Deep Neural Networks: Taxonomy and Applications](https://mlsanity.com/paper/2302.01440/) (2023)
- [OMG-ATTACK: Self-Supervised On-Manifold Generation of Transferable Evasion Attacks](https://mlsanity.com/paper/2310.03707/) (2023)
- [Mask-TS Net: Mask Temperature Scaling Uncertainty Calibration for Polyp Segmentation](https://mlsanity.com/paper/2405.05830/) (2024)
- [Controlled Paraphrase Geometry in Sentence Embedding Space: Local Manifold Modeling and Latent Probing](https://mlsanity.com/paper/2605.01073/) (2026)
- [LiBaGS: Lightweight Boundary Gap Synthesis for Targeted Synthetic Data Selection](https://mlsanity.com/paper/2605.11231/) (2026)
- [Confidence is Not Reliability: Rethinking MC Dropout in Brain Tumour Segmentation](https://mlsanity.com/paper/2606.19300/) (2026)

## Abstract (arXiv)

Uncertainty estimates help to identify ambiguous, novel, or anomalous inputs, but the reliable quantification of uncertainty has proven to be challenging for modern deep networks. In order to improve uncertainty estimation, we propose On-Manifold Adversarial Data Augmentation or OMADA, which specifically attempts to generate the most challenging examples by following an on-manifold adversarial attack path in the latent space of an autoencoder-based generative model that closely approximates decision boundaries between two or more classes. On a variety of datasets as well as on multiple diverse network architectures, OMADA consistently yields more accurate and better calibrated classifiers than baseline models, and outperforms competing approaches such as Mixup, as well as achieving similar performance to (at times better than) post-processing calibration methods such as temperature scaling. Variants of OMADA can employ different sampling schemes for ambiguous on-manifold examples based on the entropy of their estimated soft labels, which exhibit specific strengths for generalization, calibration of predicted uncertainty, or detection of out-of-distribution inputs.