AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Inherent Adversarial Robustness of Deep Spiking Neural Networks: Effects of Discrete Input Encoding and Non-Linear Activations

Saima Sharmin, Nitin Rathi, Priyadarshini Panda, Kaushik Roy

ECCV 2020 · Adversarial robustness and security › Certified and empirical robustness · approach: Robustness-accuracy trade-off analysis · 95 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
Train with surrogate-gradient backprop (SNN-BP) instead of ANN-SNN conversion onlySNN on VGG5 · CIFAR10 FGSM blackboxadversarial accuracy: 16.0 vs 9.6 (PGD), 21.5 vs 19.3 (FGSM)Improved
Use SNN-BP instead of ANNVGG5 · CIFAR10 whitebox FGSMadversarial accuracy: 10.4 -> 15Improved
Use SNN-BP instead of ANNVGG11 · CIFAR100 whitebox FGSMadversarial accuracy: 17.1 -> 15.5Hurt
Use SNN-BP instead of ANNVGG11 · CIFAR100 whitebox PGDadversarial accuracy: 8.5 -> 6.3Hurt
Use SNN-BP instead of ANNVGG11 · CIFAR100 blackbox FGSMadversarial accuracy: 21.2 -> 21.4No clear effect
Use SNN-BP instead of ANNResNet20 · CIFAR10 blackbox FGSMadversarial accuracy: 56.7 -> 56.8No clear effect
Reduce number of timesteps from 150 to 50 (4-bit input quantization) (ablation)SNN-BP VGG5 · CIFAR10 blackbox FGSMadversarial accuracy: ~5% increase; clean 88.9 -> 87.5Improved
Add 2-bit input pre-quantization instead of 4-bit (ablation)SNN-BP VGG5 · CIFAR10 blackbox FGSMadversarial accuracyImproved
Reduce number of timesteps (300 to 2000 range) (ablation)SNN-conv VGG5 · CIFAR10 blackbox FGSMadversarial accuracyNo clear effect
Add 2-bit input pre-quantization (ablation)SNN-conv VGG5 · CIFAR10 blackbox FGSMadversarial accuracy: ~21% to ~28%Improved
Increase leak (decrease lambda to 0.95) (ablation)SNN-BP VGG5 · CIFAR10 FGSM and PGDadversarial accuracy: 3~6% increaseImproved
Replace ReLU with LIF neuron on perturbed input (ablation)LIF neuron (lambda 0.99) · 32x32 image perturbationl2 norm of perturbation: 0.6 -> 0.3Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy