Institution profile

Temple University

Academic institutionnorthamerica · us
Official website
Research library132linked papers
Opportunities0open roles
Selected work

Representative Papers

HearSmoking: Smoking Detection in Driving Environment via Acoustic Sensing on Smartphones

Aug 01, 2022IEEE Transactions on Mobile Computing

Smoking while driving poses a significant threat to road safety, yet existing detection methods typically rely on intrusive sensors or auxiliary hardware. This paper proposes a contactless, end-to-end smoking behavior detection framework leveraging smartphone acoustic sensing: it exploits speaker–microphone co-design to emit and capture acoustic signals, capturing dynamic changes in acoustic correlation induced by the coupling of hand motion and thoracic respiration during smoking. We introduce the first temporal periodicity model characterizing composite smoking actions, enabling simultaneous hand-motion classification and respiratory rhythm analysis. The method integrates relative correlation coefficient computation, CNN-based feature learning, and explicit periodicity modeling. Evaluated under realistic driving conditions, it achieves real-time performance with an average accuracy of 93.44%. This work establishes a novel paradigm for unobtrusive, in-vehicle safety monitoring.

9 citationsRead paper

HearFit+: Personalized Fitness Monitoring via Audio Signals on Smart Speakers

May 01, 2023IEEE Transactions on Mobile Computing

To address the challenge of personalized, contactless fitness monitoring in home/office environments—where professional guidance and wearable devices are unavailable—this paper proposes the first smart-speaker-based acoustic sensing system for non-contact exercise monitoring. Methodologically, it integrates Doppler shift modeling, short-term energy–driven motion segmentation, and an end-to-end deep neural network, introducing a novel unified framework that jointly performs exercise action classification and user identification, with built-in incremental learning to dynamically incorporate new actions. It further defines a four-dimensional quality assessment metric encompassing duration, intensity, continuity, and fluency. Evaluated on over 9,000 repetitions of 10 exercise actions performed by 12 volunteers, the system achieves 96.13% action classification accuracy and 91% user identification accuracy, significantly enhancing autonomous training efficacy.

5 citationsRead paper

User Authentication on Earable Devices via Bone-Conducted Occlusion Sounds

Jul 01, 2024IEEE Transactions on Dependable and Secure Computing

To address the vulnerability of conventional biometrics (e.g., fingerprint, face) to spoofing attacks and the persistent trade-off between security and usability in mobile device authentication, this paper proposes TeethPass⁺—the first seamless authentication scheme leveraging bone-conducted occlusion sounds from bilateral ear canals, captured via ear-worn devices. It introduces dental occlusion–induced bone-conducted acoustics as a novel physiological biometric. Key innovations include spectrum-variance–driven event detection, time-frequency noise suppression, decoupled modeling of four-dimensional physiological traits (dental/skeletal morphology, occlusion pose, and acoustic response), and triplet-network–based embedding learning. Evaluated on 53 subjects, TeethPass⁺ achieves 98.6% authentication accuracy and 99.7% spoofing resistance, while demonstrating strong environmental robustness and zero user awareness during authentication.

3 citations1 influentialRead paper

FingerSlid: Towards Finger-Sliding Continuous Authentication on Smart Devices Via Vibration

May 01, 2024IEEE Transactions on Mobile Computing

To address the vulnerability of behavioral biometrics to spoofing attacks in continuous authentication on mobile devices, this paper proposes FingerSlid—a novel system that actively excites the device using its built-in vibration motor and captures user-specific vibrational responses induced by finger sliding via the accelerometer. Crucially, it extracts physiology-based, behavior-agnostic biometric features rather than action-dependent ones. Methodologically, FingerSlid introduces the first active-vibration-enabled finger-sliding biometric sensing paradigm, establishes a dual-modal signal acquisition pipeline, and designs a Triplet-based deep metric learning network to explicitly suppress motion-related interference—enabling truly behavior-invariant, fine-grained continuous authentication. Experimental evaluation demonstrates an average authentication accuracy of 95.4%, robust resistance to 99.5% of both synthetic and replay attacks, and strong generalizability across diverse real-world scenarios, confirming its practical viability and robustness.

2 citationsRead paper

Structure of Classifier Boundaries: Case Study for a Naive Bayes Classifier

Dec 08, 2022arXiv.org

This paper addresses the challenge of vast and structurally complex decision boundaries in DNA read mapping to reference genomes in next-generation sequencing (NGS). It investigates the boundary properties of naïve Bayes classifiers under graph-structured input spaces. To this end, the authors propose “neighborhood similarity” — a novel uncertainty measure that is both theoretically interpretable and universally computable, overcoming the reliance of conventional Bayesian confidence on model outputs. Leveraging graph-model-driven boundary analysis, neighborhood distribution statistics, and uncertainty quantification, the study reveals the high-dimensional complexity of decision boundaries and proves that the proposed measure simultaneously captures intrinsic Bayesian uncertainty. Moreover, it seamlessly extends to black-box classifiers lacking built-in confidence mechanisms. Empirically, neighborhood similarity significantly enhances classification interpretability and robustness, offering a principled framework for uncertainty-aware read mapping in NGS applications.

2 citationsRead paper
Recent publications

Latest Papers