mlsanityData to 14 Oct 2026

AERIAL: Adversarial Evaluation of Robustness in Accuracy-Preserving Low-Precision EEG Decoders

arXiv 2609.30037 ↗First posted 24 Sep 2026cs.CV cs.CR cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Deployment-oriented compression is attractive for resource-constrained brain--computer interfaces (BCIs), but whether it changes adversarial vulnerability remains unclear. On BCI Competition IV-2a, we compare 32-bit floating-point (FP32) EEGNet and ShallowConvNet models with global magnitude pruning and simulated INT8 post training quantization (PTQ) and quantization-aware training (QAT) across nine subjects and three seeds. Simulation provides differentiable quantize--dequantize models for white-box attacks and gradient analysis, while native TensorRT deployment is used for validation. Accuracy-preserving compression does not improve direct robustness: at ε=0.005, EEGNet PGD accuracy remains 22--24% across FP32, 50% pruning (P50), PTQ, and QAT. However, P50 reduces bidirectional transfer efficiency to 0.963/0.928 (FP32rightarrowP50/P50rightarrowFP32), versus 0.994/0.997 for PTQ; the same trend holds for ShallowConvNet. Gradient alignment shows a corresponding separation, while native PTQ agrees with simulated clean/adversarial predictions in 95--98% of cases. These results show that direct robustness, adversarial transfer, and deployment efficiency are distinct properties of compressed EEG decoders.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 14 of its 18 references that are in the map. First posted on arXiv 24 Sep 2026.

Builds on

14 of 18 references.

References · 14

  1. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  2. 2015Learning both Weights and Connections for Efficient Neural Networkscited by 2,031
  3. 2017Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inferencecited by 1,593
  4. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,268
  5. 2016EEGNet: A Compact Convolutional Network for EEG-based Brain-Computer Interfacescited by 628
  6. 2021A White Paper on Neural Network Quantizationcited by 472
  7. 2019On Evaluating Adversarial Robustnesscited by 447
  8. 2019On the Vulnerability of CNN Classifiers in EEG-Based BCIscited by 17
  9. 2020Improved Gradient based Adversarial Attacks for Quantized Networkscited by 10
  10. 2024Adversarial Pruning: A Survey and Benchmark of Pruning Methods for Adversarial Robustnesscited by 10
  11. 2019Impact of Low-bitwidth Quantization on the Adversarial Robustness for Embedded Neural Networkscited by 9
  12. 2020Q-EEGNet: an Energy-Efficient 8-bit Quantized Parallel EEGNet Implementation for Edge Motor-Imagery Brain--Machine Interfacescited by 8
2 more
  1. 2022MI-BMInet: An Efficient Convolutional Neural Network for Motor Imagery Brain--Machine Interfaces with EEG Channel Selectioncited by 7
  2. 2024Alignment-Based Adversarial Training (ABAT) for Improving the Robustness and Accuracy of EEG-Based BCIscited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Pretrained, Frozen, Still Leaking: Auditing Cross-Encoder Attribute Transfer in EEG Foundation Models
  2. 2021On the Adversarial Robustness of Quantized Neural Networks
  3. 2026What masking geometry works best for EEG foundation models?
  4. 2025Reducing Storage of Pretrained Neural Networks by Rate-Constrained Quantization and Entropy Coding
  5. 2026BrainRVQ: A High-Fidelity EEG Foundation Model via Dual-Domain Residual Quantization and Hierarchical Autoregression
  6. 2026EEG-Fusion: Failure-Informed Source-Free Expert Routing for Robust Motor Imagery EEG Decoding
  7. 2022Optimal Brain Compression: A Framework for Accurate Post-Training Quantization and Pruning
  8. 2026BrainDistill: Implantable Motor Decoding with Task-Specific Knowledge Distillation