Defense Notices

Person re-identification (Re-ID) and multi-object tracking in unconstrained surveillance environments pose significant challenges within the field of computer vision. These complexities stem mainly from occlusion, variability in appearance, and identity switching across various camera views. This research outlines a comprehensive and innovative agenda aimed at tackling these issues, employing a series of increasingly advanced deep learning architectures, culminating in a groundbreaking occlusion-aware Vision Transformer framework.

At the heart of this work is the introduction of Deep SORT with Multiple Inputs (Deep SORT-MI), a cutting-edge real-time Re-ID system featuring a dual-metric association strategy. This strategy adeptly combines Mahalanobis distance for motion-based tracking with cosine similarity for appearance-based re-identification. As a result, this method significantly decreases identity switching compared to the baseline SORT algorithm on the MOT-16 benchmark, thereby establishing a robust foundation for metric learning in subsequent research.

Expanding on this foundation, a novel pose-estimation framework integrates 2D skeletal keypoint features extracted via OpenPose directly into the association pipeline. By capturing the spatial relationships among body joints along with appearance features, this system enhances robustness against posture variations and partial occlusion. Consequently, it achieves substantial reductions in false positives and identity switches compared to earlier methods, showcasing its practical viability.

Furthermore, a Diverse Detector Integration (DDI) study meticulously assessed the influence of detector choices—including YOLO v4, Faster R-CNN, MobileNet SSD v2, and Deep SORT—on the efficacy of metric learning-based tracking. The results reveal that YOLO v4 consistently delivers exceptional tracking accuracy on both the MOT-16 and MOT-17 datasets, establishing its superiority in this competitive landscape.

In conclusion, this body of research notably advances occlusion-aware person Re-ID by illustrating a clear progression from metric learning to pose-guided feature extraction and ultimately to transformer-based global attention modeling. The findings underscore that lightweight, meticulously parameterized Vision Transformers can achieve impressive generalization for occlusion detection, even under constrained data scenarios. This opens up exciting prospects for integrated detection, localization, and re-identification in real-world surveillance systems, promising to enhance their effectiveness and reliability.

Android malware remains a significant threat to mobile security, requiring efficient and scalable detection methods. This project presents an Android Malware Detection System that uses machine learning to classify applications as benign or malicious based on static permission-based analysis. The system is trained on the TUANDROMD dataset of 4,464 applications using four models-Logistic Regression, XGBoost, Random Forest, and Naive Bayes-with a 75/25 train/test split and 5-fold cross-validation on the training set for evaluation. To improve reliability, the system incorporates a hybrid decision approach that combines machine learning confidence scores with a rule-based static analysis engine, using a three-zone confidence routing mechanism to capture threats that ML alone may miss. The solution is deployed as a Flask web application with both a manual detection interface and an APK file scanner, providing predictions, confidence scores, and risk insights, ultimately supporting more informed and secure decision-making.

Architectural design space exploration (DSE) requires navigating large configuration spaces while satisfying multiple conflicting objectives, including performance, cost, and system dependability. Large language models (LLMs) have shown promise in assisting DSE by proposing candidate designs and interpreting simulation feedback. However, extending LLM-based DSE to realistic multi-objective settings introduces structural challenges. A naive multi-objective extension of prior LLM-based DSE approaches, which we term Co-Pilot2, exhibits reasoning instability, candidate degeneration, feasibility violations, and lack of progressive improvement. These limitations arise not from insufficient model capacity, but from the absence of structured control, verification, and decision integrity within the exploration process. 

To address these challenges, this research introduces REMODEL, a structured LLM-controlled DSE framework that transforms free-form reasoning into a constrained, verifiable, and iterative optimization process. REMODEL incorporates candidate pooling across parallel reasoning instances, strict state isolation via history snapshotting, deterministic feasibility verification, canonical design representation and deduplication, explicit decision stages, and structured reasoning to enforce complete parameter coverage and consistent trend analysis. These mechanisms enable reliable and stable exploration under complex multi-objective constraints. 

To support dependability-aware evaluation, the framework is integrated with cycle-accurate simulation using gem5 and its reliability-focused extension GemV, enabling detailed analysis of performance, power, and fault tolerance through vulnerability metrics. This integration allows the system to reason not only about performance–cost trade-offs, but also about reliability-aware design decisions under realistic execution conditions. 

Experimental evaluation demonstrates that REMODEL identifies near-optimal designs within a small number of simulations, achieving significantly higher solution quality per simulation compared to baseline methods such as random search and genetic algorithms, while maintaining low computational overhead. 

This work establishes a foundation for dependable LLM-assisted DSE by incorporating reliability constraints into the exploration loop. As a future direction, this framework will be extended to incorporate security-aware design considerations, enabling unified reasoning over performance, cost, reliability, and system security. 

This thesis addresses two related problems in the optical manipulation of microscopic particles: the efficient generation of holograms for holographic optical tweezers and the simulation of multi-particle optical binding. Holographic optical tweezers use phase-only spatial light modulators to create programmable optical trapping fields, enabling dynamic control over the number, position, and relative strength of optical traps. Because the quality of the trapping field depends strongly on the computed hologram, the first part of this work focuses on improving hologram-generation methods used in these systems.

A new phase-induced compressive sensing algorithm is presented for holographic optical tweezers, along with weighted and unweighted variants. These methods are developed from the Gerchberg-Saxton framework and are designed to improve computational efficiency while preserving favorable trapping characteristics such as uniformity and optical efficiency. By combining compressive sensing with phase induction, the proposed algorithms reduce the computational burden associated with iterative hologram generation while maintaining strong performance across a variety of trapping arrangements. Comparative simulations are used to evaluate these methods against several established hologram-generation algorithms, and the results show that the proposed approaches offer meaningful improvements in convergence behavior and overall performance.

The second part of this thesis examines optical binding, a phenomenon in which multiple particles interact through both the incident optical field and the fields scattered by neighboring particles. To study this process, a numerical simulation is developed that incorporates gradient forces, radiation pressure, and light-mediated particle-particle interactions in both two- and three-dimensional configurations. The simulation is used to investigate how particles evolve under different initial conditions and illumination states, and how collective effects influence the formation of stable or semi-stable arrangements. These results provide insight into the role of scattering-mediated forces in many-particle optical systems and highlight differences between two-dimensional and three-dimensional behavior.

Although hologram generation and optical binding are treated as separate problems in this work, they are connected by a common goal: understanding how structured optical fields can be designed and applied to control microscopic matter. Together, the results of this thesis contribute to the broader study of computational beam shaping and many-body optical interactions, with relevance to advanced optical trapping, particle organization, and dynamically reconfigurable light-driven systems.

Runtime estimation plays a structural role in reservation-based scheduling for High Performance Computing (HPC) systems, where predicted walltimes directly influence reservation timing, backfilling feasibility, and overall queue dynamics. This raises a fundamental question of whether improved runtime prediction accuracy necessarily translates into improved scheduling performance. In this work, we conduct an empirical study of runtime estimation under EASY Backfilling using an application-driven workload consisting of MRI-based brain segmentation jobs. Despite identical configurations and uniform metadata, runtimes exhibit substantial variability driven by intrinsic input structure. To capture this variability, we develop a feature-driven machine learning (ML) framework that extracts region-wise features from MRI volumes to predict job runtimes without relying on historical execution traces or scheduling metadata. We integrate these ML-derived predictions into an EASY Backfilling scheduler implemented in the Batsim simulation framework. Our results show that regression accuracy alone does not determine scheduling performance. Instead, scheduling performance depends strongly on estimation bias and its effect on reservation timing and runtime exceedances. In particular, mild multiplicative calibration of ML-based runtime estimates stabilizes scheduler behavior and yields consistently competitive performance across workload and system configurations. Comparable performance can also be observed with certain levels of uniform overestimation; however, calibrated ML predictions provide a systematic mechanism to control estimation bias without relying on arbitrary static inflation. In contrast, underestimation consistently leads to severe performance degradation and cascading job terminations. These findings highlight runtime estimation as a structural control input in backfilling-based HPC scheduling and demonstrate the importance of evaluating prediction models jointly with scheduling dynamics rather than through regression metrics alone.

The University of Kansas maintains hundreds of departmental and unit websites, leaving students without a unified way to find information. General-purpose chatbots hallucinate KU-specific facts, and static FAQ pages cannot hold a conversation. This work presents BabyJay, a Retrieval-Augmented Generation chatbot that answers student questions using content scraped from official KU sources, with inline citations on every response. The pipeline combines query preprocessing and decomposition, an intent classifier that routes most queries to fast JSON lookups, hybrid retrieval (BM25 and ChromaDB vector search merged via Reciprocal Rank Fusion), a cross-encoder re-ranker, and generation by Claude Sonnet 4.6 under a context-only system prompt. Evaluation on 46 question-answer pairs across five difficulty tiers and eight domains produced a composite score of 0.72, entity precision of 93%, and zero runtime errors. Retrieval, rather than generation, emerged as the primary bottleneck, motivating future work on multi-domain query handling.

Modern intelligent systems increasingly rely on continuous sensor data streams for perception, decision-making, and control, making sensors a critical yet underexplored attack surface. While prior research has demonstrated the feasibility of sensor-based attacks, recent advances in mobile operating systems and machine learning-based defenses have significantly reduced their practicality, rendering them more detectable, resource-intensive, and constrained by evolving permission and context-aware security models.

This dissertation revisits sensor exploitation under these modern constraints and develops a unified, cross-layer perspective that improves both practicality and stealth of sensor-enabled attacks. We identify three fundamental challenges: (i) the difficulty of reliably manipulating physical sensor signals in noisy, real-world environments; (ii) the effectiveness of context-aware defenses in detecting anomalous sensor behavior on mobile devices, and (iii) the lack of lightweight coordination for practical sensor-based side- and covert-channels.

To address the first challenge, we propose a physical-domain attack framework that integrates signal modeling, simulation-guided attack synthesis, and real-time adaptive targeting, enabling robust adversarial perturbations with high attack success rates even under environmental uncertainty. As a case study, we demonstrate an infrared laser-based adversarial example attack against face recognition systems, which achieves consistently high success rates across diverse conditions with practical execution overhead.

To improve attack stealth against context-aware defenses, we introduce an auto-contextualization mechanism that synchronizes malicious sensor actuation with legitimate application activity. By aligning injected signals with both statistical patterns and semantic context of benign behavior, the approach renders attacks indistinguishable from normal system operations and benign sensor usage. We validate this design using three Android logic bombs, showing that auto-contextualized triggers can evade both rule-based and learning-based detection mechanisms.

Finally, we extend sensor exploitation beyond the traditional attack-channel plane by introducing a lightweight control-plane protocol embedded within sensor data streams. This protocol encodes control signals directly into sensor observations and leverages simple signal-processing primitives to coordinate multi-stage attacks without relying on privileged APls or explicit inter-process communication. The resulting design enables low-overhead, stealthy coordination of cross-device side- and covert-channels.

Together, these contributions establish a new paradigm for sensor exploitation that spans physical, contextual, and control-plane dimensions. By bridging these layers, this dissertation demonstrates that sensor-based attacks remain not only feasible but also practical and stealthy in modern computer systems.

Modern phased-array antenna calibration is essential for advanced radar systems to achieve precise beamforming, sidelobe control, and coherent processing. While mutual coupling-based calibration provides a valuable internal alternative to external far-field references by exploiting near-field element interactions, the problem is fundamentally ill-posed. Measured responses depend simultaneously on transmit coefficients, receive coefficients, and the coupling matrix, making it difficult to isolate true channel errors from array-model mismatch without additional structure.

This thesis presents a Bayesian Maximum A Posteriori (MAP) calibration framework that resolves this ambiguity by embedding physically motivated prior information into the estimation problem. The nominal coupling matrix is decomposed into Infinite, Symmetric, and Reciprocal components, which define low-dimensional parameterizations and prior covariance models. A Maximum Likelihood (ML) stage first generates a data-consistent transceiver initialization, followed by a MAP estimator that refines the solution by jointly addressing structured coupling deviations and measurement uncertainty.

Evaluations using Computational Electromagnetic (CEM) models and measured WaDES array data reveal that the physical array contains more higher-order structural content than the nominal CEM model. Across Monte Carlo trials, highly structured MAP estimators generally achieve lower aggregate error than unconstrained ML and Log Least Squares (LLS) methods. The overlapping-subspace M family offers an optimal balance of structural flexibility, zero-centered phase and magnitude behavior, and tuning robustness. Additionally, parametric sweeps highlight that prior covariance scaling is a critical design parameter: tight reciprocal priors prevent spurious structural absorption, whereas overly loose priors allow model mismatch to contaminate transceiver estimates.

Ultimately, this work demonstrates that internal mutual coupling calibration can achieve autonomy and robustness against model mismatch by parameterizing the nominal coupling matrix into structured components and integrating them as Bayesian priors.

Substance use remains a major public health issue in the United States that significantly impacts individuals, families, and society. Many individuals who suffer from substance use disorder (SUD) face incarceration due to drug-related offenses. Drug courts have emerged as an alternative to imprisonment and offer the opportunity for individuals to participate in a drug rehabilitation program instead. Drug courts mainly focus on those with non-violent drug-related offenses. One of the challenges of decision making in drug courts is assessing the likelihood of participants graduating from the drug court and avoiding recidivism after graduation. This study investigates the use of machine learning models to predict success in drug courts using data from a substance use drug court in Missouri. Success is measured in terms of graduation from the program, and the model includes a wide range of potential predictors including demographic characteristics, family and social factors, substance use history, legal involvement, physical and mental health history, employment history as well as drug court participation data. The results will be beneficial to drug court teams and presiding judges in predicting client success, evaluating risk factors during treatment for participants, informing person-centered treatment planning, and the development of after-care plans for high-risk participants to reduce the likelihood of recidivism. 

Optical Neural Networks (ONNs) have gained traction as an alternative to the conventional computing architectures used in modern CPUs and GPUs, largely because light enables massive parallelism, ultrafast inference, and minimal power consumption. 

As with conventional deep neural networks (DNNs), free-space ONNs require two main layers: (1) a nonlinear activation function which exists to separate adjacent linear layers, and (2) weighting layers that applies a linear transformation given an input.

Firstly, a Resonant Semiconductor Saturable Absorption Mirror (RSAM) was investigated as a viable nonlinear activation function. Several mechanisms have been used to create nonlinear activation functions, such as cold atoms, vapor absorption cells, and polaritons, but these implementations are bulky and must operate under tightly controlled environments while RSAMs is a passive device. Compared to typical SESAMs, the resonance structure of RSAM also reduces the saturation fluence compared to non-resonant SAMs, allowing low power laser sources to be used. A fiber-based optical testbed demonstrated notable improvement of 8.1% in classification accuracy compared to a linear only network trained with the MNIST dataset.

Secondly, Micro-electromechanical-system-based phase light modulators (PLMs) were evaluated as an alternative to LC-SLMs for in-situ reinforcement learning. PLMs can operate at kilohertz-scale frame rates at a substantially lower cost compared to LC-SLMs but have lower phase resolution and non-uniform quantization which impacts fidelity. Despite these disadvantages, the high-speed nature of PLMs allows for significant decrease in optimization time, which not only allows for reduction in training time, but also allows for larger datasets and more complex models with more learnable parameters. A single layer optical network was implemented using policy-based learning with discrete action-space to minimize impact of quantization. The testbed achieves 90.1%, 79.7%, and 76.9% training, validation, and test accuracy, respectively, on 3,000 images from the MNIST dataset. Additionally, we achieved 79.9%, 72.1%, and 71.7% accuracy on 3,000 images from the Fashion MNIST dataset. At 14 minutes per epoch during training, it is at least a magnitude lower in training time compared to LC-SLMs based models.