Search

This article investigates the potential of surface electromyography (sEMG) as a new biometric modality and proposes a deep neural network architecture as the backbone of a gesture-independent personal identification system (PIS). This article focuses on the real-world translation of such a model through systematic optimization, which finds the minimum number of gestures and sensors needed for training. Focusing on “dynamic sEMG,” our proposed method can successfully identify 40 subjects with an average accuracy of 97%. This is achieved when gestures are the same in training, validation, and testing (the subjects need to repeat a particular gesture among a set of seven known gestures as a passcode). In a more complex scenario, when training gestures differ from those in validation and testing, our model can achieve an average accuracy of 90%, demonstrating that the proposed model can extract the unique patterns to identify a user regardless of gestures. Taking advantage of gradient-weighted class activation mapping (Grad-CAM), we explore the attention of the model on segments of the spectrotemporal space of the input signals. Grad-CAM not only sheds light on sEMG-based personal identification by decoding and visualizing the unique user-specific neurophysiological pattern but also generates a 2-D spectrotemporal mask used to reduce the model complexity significantly. As a result of the systematic optimization and Grad-CAM analysis, our proposed identification method needs only 4% of data for training, boosting practicality. This article also reveals the robustness of the proposed model for cross-day evaluation. Finally, the comparative study shows the superiority of our proposed model over several state-of-the-art algorithms.

We propose a framework in which multiple entities collaborate to build a machine learning model while preserving privacy of their data. The approach utilizes feature embeddings from shared/per-entity feature extractors transforming data into a feature space for cooperation between entities. We propose two specific methods and compare them with a baseline method. In Shared Feature Extractor (SFE) Learning, the entities use a shared feature extractor to compute feature embeddings of samples. In Locally Trained Feature Extractor (LTFE) Learning, each entity uses a separate feature extractor, and models are trained using concatenated features from all entities. As a baseline, in Cooperatively Trained Feature Extractor (CTFE) Learning, the entities train models by sharing raw data. Secure multi-party algorithms are utilized to train models without revealing data or features in plain text. We investigate the trade-offs among SFE, LTFE, and CTFE in regard to performance, privacy leakage (using an off-the-shelf membership inference attack), and computational cost. LTFE provides the most privacy, followed by SFE, and then CTFE. Computational cost is lowest for SFE and the relative speed of CTFE and LTFE depends on network architecture. CTFE and LTFE provide the best accuracy. We use three different datasets for evaluations.

This study explores data-driven detection of firmware/software Trojans in embedded systems without golden models. We consider embedded systems such as single board computers and industrial controllers. While prior literature considers side channel based anomaly detection, this study addresses the following central question: is anomaly detection feasible when using low-fidelity simulated data without using data from a known-good (golden) system? To study this question, we use data from a simulator-based proxy as a stand-in for unavailable golden data from a known-good system. Using data generated from the simulator, one-class classifier machine learning models are applied to detect discrepancies against expected side channel signal patterns and their inter-relationships. Side channels fused for Trojan detection include multi-modal side channel measurement data (such as Hardware Performance Counters, processor load, temperature, and power consumption). Additionally, fuzzing is introduced to increase detectability of Trojans. To experimentally evaluate the approach, we generate low-fidelity data using a simulator implemented with a component-based model and an information bottleneck based on Gaussian stochastic models. We consider example Trojans and show that fuzzing-aided golden-free Trojan detection is feasible using simulated data as a baseline.