# Improving Robustness of Spectrogram Classifiers with Neural Stochastic Differential Equations

arXiv: https://arxiv.org/abs/2409.01532 · HTML: https://mlsanity.com/paper/2409.01532/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 30 references that are in the map.
- First posted on arXiv 3 Sep 2024.

## Builds on (most cited first, 11 of 11)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [SmoothGrad: removing noise by adding noise](https://mlsanity.com/paper/1706.03825/) (2017)
- [ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders](https://mlsanity.com/paper/2301.00808/) (2023)
- [Captum: A unified and generic model interpretability library for PyTorch](https://mlsanity.com/paper/2009.07896/) (2020)
- [Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations](https://mlsanity.com/paper/1710.10121/) (2017)
- [Neural SDE: Stabilizing Neural ODE Networks with Stochastic Noise](https://mlsanity.com/paper/1906.02355/) (2019)
- [On learning optimized reaction diffusion processes for effective image restoration](https://mlsanity.com/paper/1503.05768/) (2015)
- [ResNets Ensemble via the Feynman-Kac Formalism to Improve Natural and Robust Accuracies](https://mlsanity.com/paper/1811.10745/) (2018)
- [Improved Speech Emotion Recognition using Transfer Learning and Spectrogram Augmentation](https://mlsanity.com/paper/2108.02510/) (2021)
- [Explanation of Unintended Radiated Emission Classification via LIME](https://mlsanity.com/paper/2009.02418/) (2020)
- [Explainable Signature-based Machine Learning Approach for Identification of Faults in Grid-Connected Photovoltaic Systems](https://mlsanity.com/paper/2112.14842/) (2021)

## Abstract (arXiv)

Signal analysis and classification is fraught with high levels of noise and perturbation. Computer-vision-based deep learning models applied to spectrograms have proven useful in the field of signal classification and detection; however, these methods aren't designed to handle the low signal-to-noise ratios inherent within non-vision signal processing tasks. While they are powerful, they are currently not the method of choice in the inherently noisy and dynamic critical infrastructure domain, such as smart-grid sensing, anomaly detection, and non-intrusive load monitoring.