arXiv subjects

Phuc Nguyen

Publications and source records attributed to Phuc Nguyen.

At least 19 recordsLinked to original sources

arXiv2026

Real-Time Recurrent Qubit-State Discrimination on a Heterogeneous AI-Engine-FPGA Accelerator

Qubit readout, or qubit state discrimination, remains one of the slowest and most error-prone operations in superconducting quantum processors, making fast, accurate, and hardware-efficient classification essential for scalable quantum computing. Conventional discriminators integrate the I/Q record into a single point, discarding evidence a relaxing qubit leaves behind; full-resolution feed-forward classifiers recover it but are too costly to deploy under this constraint. Recurrent classifiers retain full resolution far more cheaply, but their loop-carried hidden state cannot fit a single AI Engine (AIE) tile's register capacity and spills to data memory, adding latency that grows with sequence length. We present a hardware-aware LSTM accelerator for an AMD Versal device that resolves this by partitioning at the dependency boundary - independent input projections on the AIE array, the loop-carried update in programmable logic (PL) - keeping pace with a 64 ns input cadence under a 1,000 ns budget. On the qubit affected by mid-readout state transitions, our design matches a full-resolution feed-forward network's accuracy (96.59% vs. 96.37%) using 403x fewer parameters and roughly 103x fewer DSPs while cutting relative error by 68.2% over GMM, 43.6% over QubiCML, and 16.82% over MF-NN. Using 4 out of 400 AIE tiles and 17 DSP58, it needs 2.2-8.2x fewer DSP58 blocks than QubiCML and MF-NN, and the AIE-PL partition cuts exposed tail latency by 60.3% versus a pure-AIE mapping and DSP58 usage 31.1x versus an all-PL implementation. Across eight qubits, a compact variant reaches 96.78% mean accuracy, the strongest hardware-deployed operating point. Temporal resolution and recurrent-state placement must then be treated jointly when designing low-latency quantum readout accelerators.

cs.AR↗

arXiv2026

HyVox: Rethinking Low-Power Voice Intelligence towards Neuromorphic-von Neumann Edge Computing

We present HyVox, a hybrid neuromorphic--von Neumann system for low-energy open-set speaker identification under tight energy budgets. HyVox splits the workload at the speaker embedding. A neuromorphic processor embeds the speech with sparse, event-driven computation, and a von Neumann microcontroller holds the speaker gallery and makes every decision: it scores queries by cosine similarity, rejects unknown speakers, and enrolls new users by writing a prototype, with no weight update. On the neuromorphic processor, a hierarchical temporal abstraction (HTA) SNN backbone learns a synaptic time constant per leaky integrate-and-fire (LIF) neuron, allocating 832 neurons across fast, medium, and slow timescales. A 32-neuron spiking embedding extractor (SEE), pretrained on episodes that each enroll new speakers and reject outsiders, turns the backbone output into an embedding for speakers unseen in training; its 32 spike counts are the only interface between the two processors. To fit the chip's synapse and integer budgets, we compress the backbone 6.5x to 29.0k synapses, train it for integer bit-shift dynamics, and fine-tune it with chip-accurate spiking training (CAST). CAST trains through a differentiable replica of the chip's integer neuron update, derived from its cycle-accurate simulator, so that gradient descent optimizes the exact weights and thresholds the chip runs; it doubles the top-1 accuracy of the mapped backbone on unseen speakers. On 20 Common Voice households of unseen speakers, HyVox reaches 99.6\% top-1 and 5.5\% rejection EER with 2 members and 94.9\% and 13.2\% with 20, and the network deployed on the neuromorphic chip matches its trained model. Each inference on a 3-s window takes 0.96 s and 1.63 mJ of dynamic energy, 42\% lower latency than the fastest embedded ANN baseline and 13.5x less dynamic and 5.8x less total energy than the most energy-efficient one.

cs.NE↗

arXiv2026

SeRV: Semantic-Aligned Residual Vector Quantization for American Sign Language Generation

American Sign Language (ASL) generation remains challenging due to limited paired text-ASL motion data and the difficulty of learning motion representations both precise for reconstruction and predictable from linguistic input. Existing methods rely on motion tokenizers optimized for reconstruction, without explicit semantic supervision from paired text. As a result, the learned tokens remain limited in supporting semantically consistent and fine-grained ASL motion generation. To address this limitation, we propose SeRV (Semantic-Aligned Residual Vector Quantization), a semantic-aligned RVQ tokenizer for ASL generation. SeRV learns a semantically structured residual token space by combining global-level motion-text alignment with token-level text-conditioned supervision. Building on this tokenizer, a Hierarchical GPT predicts residual motion tokens in a coarse-to-fine manner, generating structurally coherent and semantically aligned 3D ASL motion. We further construct a large-scale reconstructed 3D ASL motion-text benchmark by recovering paired 3D motion from OpenASL videos. Experiments on How2Sign and OpenASL show that SeRV outperforms six state-of-the-art baselines, with user studies assessing reconstruction quality and semantic fidelity. Samples of our reconstructed 3D ASL motion are available on our project page: https://signvideos.github.io/Demo/.

cs.CV↗

arXiv2026

EMG-FM-Bench: A Comprehensive Benchmark for Foundation Model Transfer and Adaptation on Electromyography

Foundation models (FMs) are increasingly being developed for general time series and physiological signals, yet their transferability to downstream physiological tasks remains poorly understood. This question is particularly challenging for electromyography (EMG), where signal distributions vary substantially across users, sensing configurations, acquisition hardware, and downstream tasks. We introduce EMG-FM-Bench, a systematic benchmark for studying foundation-model transfer and adaptation on EMG. EMG-FM-Bench unifies 20 public datasets with over 1 million EMG segments and evaluates nine pretrained foundation models across four questions: how pretrained models perform when frozen or fully fine-tuned, how much pretraining helps compared with training the same model from scratch, how well models generalize to new users with limited labeled data, and how performance changes across different EMG tasks. Across the benchmark, linear probing provides useful information about pretrained representations, but full fine-tuning can substantially change downstream EMG performance. Comparing each pretrained model with the same model trained from scratch shows that the benefit of pretraining varies substantially across models and is not universal. Performance decreases when models are evaluated on new users, while five-shot adaptation improves macro-F1 in 70.2% of evaluated model-dataset combinations but recovers only part of the lost performance. Model performance is highly consistent between upper- and lower-limb classification and remains strongly correlated with continuous EMG-to-text decoding. Together, these results provide a systematic view of when pretrained time-series models transfer effectively to EMG and how their performance depends on fine-tuning, user variation, and downstream task.

cs.LG↗

arXiv2026

LiteEMG-FM: An Efficient and Deployable Foundation Model for Robust EMG Sensing

Electromyography (EMG) signals vary substantially across individuals, body regions, recording sessions, and sensing hardware, limiting the generalization of models for assistive devices and human-computer interaction. Existing time-series foundation models are also computationally expensive for real-time wearable deployment and often fail to capture EMG-specific time-frequency characteristics. We present LiteEMG-FM, an efficient hybrid CNN-Transformer foundation model for practical EMG sensing. Pretrained on 16 diverse upper- and lower-limb EMG datasets, LiteEMG-FM learns representations that generalize across users and datasets. For resource-constrained deployment, we implement a hierarchical wake-up architecture in which a lightweight, always-on 1D-CNN filters rest and non-target activity and activates LiteEMG-FM only for valid gestures. We evaluate full inference offloading, split inference, and full on-device processing, characterizing their trade-offs in latency, power consumption, and memory footprint. Across diverse evaluation settings, LiteEMG-FM outperforms state-of-the-art time-series foundation models and supervised baselines, particularly under zero-calibration cross-participant and data-scarce conditions. These results demonstrate that LiteEMG-FM is an effective, efficient, and deployable foundation model for EMG applications.

cs.LG↗

arXiv2026

E3Sense: Head-Confined Multimodal Sensing of Learner Engagement

Engagement-aware learning systems could provide hints or adjust pacing when learners struggle. Prior engagement sensing work distributes sensors across or outside the body rather than consolidating them at one site, or reduces engagement to shared affect or a single dimension (from behavioral, emotional, and cognitive engagement). We introduce E3Sense, a head-worn platform that co-locates electroencephalography, eye tracking, and electrodermal activity to personalize engagement measurement. During a lab study we collected 450 ratings of engagement levels on a five-level ordinal scale while participants watched educational videos. For fifteen held-out participants, E3Sense achieved a within-one-level prediction score of 75.0%, compared with 63.0% for always predicting the most common rating. In an exploratory analysis of 18 participants from the same study who defined engagement, conditioning on learners' definitions raised the same measure by 6.9 points, from 64.6% to 71.5%. Our work provides a proof-of-concept of a head-site, personalized multimodal sensing of engagement for adaptive educational interfaces.

cs.HC↗

arXiv2026

PanoSeg3R: Feed-Forward 3D Semantic Segmentation for Panoramic Images with an Automatic Data Curation Pipeline

We present PanoSeg3R, a feed-forward framework for 3D panoramic semantic segmentation. Unlike existing methods designed for perspective inputs, PanoSeg3R jointly predicts 3D geometry and multi-view semantic segmentation in one single forward pass. Built upon a pretrained reconstruction backbone that supports panoramic images, our approach extends feed-forward 3D reconstruction with a query-based mask decoder. Furthermore, we introduce an automatic panorama data curation pipeline that leverages the complementary strengths of off-the-shelf foundation models to generate reliable pseudo semantic annotations, substantially expanding the training data and improving zero-shot generalization. PanoSeg3R achieves state-of-the-art performance on panoramic 3D semantic segmentation, improving 3D mIoU by up to 16.02 on ScanNet++, while the curated training data further improves zero-shot performance by up to 4.26 and 43.28 mIoU on Stanford2D3D and ToF-360, respectively. Website: https://harryyoon777.github.io/PanoSeg3R/

cs.CV↗

arXiv2026

ML-Powered FPGA-based Real-Time Quantum State Discrimination Enabling Mid-circuit Measurements

Accurate and timely quantum state discrimination is critical for quantum computing, particularly in protocols requiring mid-circuit measurement (MCM) and conditional feed-forward. While classical state readout in transistors is near-instantaneous, identifying the state of superconducting qubits remains latency-limited and error-prone. Existing approaches rely on post-processing measurement data transferred to host computers, introducing delays on the order of milliseconds -- vastly exceeding the coherence time of quantum states, which typically last only hundreds of microseconds. To bridge this latency gap, we present an in-situ machine learning (ML) inference engine implemented on a field-programmable gate array (FPGA), enabling real-time quantum state discrimination within 40 ns. Our system supports both two-level (qubit) and three-level (qutrit) systems and performs inference directly on digitized readout signals without host-side intervention. This low-latency operation facilitates mid-circuit measurements, where qubit states are measured and reused within a single quantum circuit, and enables real-time conditional operations essential for quantum error correction. Crucially, fast and accurate mid-circuit readout is a prerequisite for implementing practical fault-tolerant error correction, where errors must be detected and corrected within a single circuit cycle to preserve logical state coherence. We validate our system using superconducting transmon devices, demonstrating robust discrimination fidelity across multiple qubit and qutrit channels. Further, we implement a conditional qutrit logic protocol based on FPGA-resident state classification, highlighting the practical benefits of our approach for NISQ-era quantum algorithms and scalable fault-tolerant architectures.

quant-ph↗

arXiv2026

From Prediction to Incrementality: Causal Optimization for Large-Scale Targeting and Recommendation

Large-scale targeting and recommendation systems are typically built around predictive scores fed into heuristic or local allocation. When the business goal is incremental impact, as in marketing campaigns, incentives, and notifications, this paradigm systematically misallocates resources toward users who would have acted anyway. We present a decision-centric framework that instead optimizes causal effects under global constraints, aligning three components under a single objective: a causal neural network with a Transformer backbone for individual treatment-effect estimation, a Bayesian neural-bandit layer for uncertainty-aware exploration, and a dual-based large-scale linear-programming layer for constrained allocation. The framework also supports sequential context and multi-outcome, attribute-conditioned scoring through a Transformer encoder and outcome embeddings. We evaluate it with offline simulations on a public bandit dataset, targeted architectural ablations, and an online A/B test on LinkedIn Feed marketing traffic. We also distill production lessons on causal training-data construction and cost and delivery control, which were critical to successful deployment. The end-to-end treatment policy delivered a statistically significant $+7.20\%$ lift in the primary long-term-value metric, demonstrating the feasibility of production-scale causal optimization under business constraints.

cs.LG↗

arXiv2026

Transformation-Augmented GRPO for Enhancing Exploration in Reasoning of Large Language Models

Group Relative Policy Optimization (GRPO) has become the dominant method for reinforcement learning with verifiable rewards in large language models, but it suffers from two critical limitations: gradient vanishing and diversity collapse. When training questions are too easy or too hard, all sampled responses receive identical rewards, yielding zero gradients. Meanwhile, the model tends to collapse its responses toward a single reasoning pattern rather than exploring diverse strategies. We propose Transformation-Augmented GRPO (TA-GRPO), a simple but effective method that addresses both issues via question rephrasing. For each training question, we automatically generate multiple problem-equivalent rephrasings that alter wording, format, and information order while preserving the underlying meaning. Because these rephrasings shift the model's perceived difficulty, pooling responses across the original and its rephrasings yields mixed rewards and more diverse reasoning paths. TA-GRPO jointly computes advantages over this expanded response set and aligns all importance ratios to the original question, enabling the model to learn from a richer set of solution attempts. Experiments on four LLMs (Qwen3-1.7B, Qwen3-4B, Llama-3.2-1B, Llama-3.2-3B) show that TA-GRPO consistently improves pass@$k$ on competition-level benchmarks (AMC, OlympiadBench, AIME24, AIME25) and out-of-distribution benchmarks (Minerva, GPQA-Diamond). Notably, it improves the average pass@32 of Qwen3-1.7B and Qwen3-4B by \textbf{4.97} and \textbf{4.34} points, respectively, and matches the exploration quality of baselines trained on up to 2.5$\times$ more data.

cs.LG↗

arXiv2026

GEM: Gear-based Environment-Integrated Mobility for Adaptive Indoor Human Sensing

Infrastructure-based sensing systems, like Wi-Fi, thermal, vibration-based approaches, provide continuous and unobtrusive indoor human monitoring services. They are often deployed statically for long-term continuous monitoring, which often leads to inefficient sensing/inflexible deployment due to human mobility or high maintenance/data volume for dense deployments. In contrast, autonomous and human carried mobile devices can better adapt to human mobility. However, their physical presence (e.g., drones or robots) may induce observer effects, while their operation often imposes additional burdens, such as wearing (e.g., wearables) and frequent charging. We present GEM, a hybrid scheme that introduces the mobility to infrastructure-based sensing. GEM integrates a matrix of gears into everyday surfaces (e.g., floors, walls) to turn them into "public transportation" for moving infrastructure sensors around. We design and fabricate a 3 x 3 gear matrix prototype that can effectively move sensors from one location to another. We further validate the scalability of the design through simulation of up to 64 x 64 gear matrix with concurrent sensors.

eess.SY↗

arXiv2026

Which Similarity-Sensitive Entropy (Sentropy)?

Shannon entropy is not the only entropy that is relevant to machine-learning datasets, nor possibly even the most important one. Traditional entropies such as Shannon entropy capture information represented by elements' frequencies but not the richer information encoded by their similarities and differences. Capturing the latter requires similarity-sensitive entropy (``sentropy''). Sentropy can be measured using either the recently developed Leinster-Cobbold-Reeve framework (LCR) or the newer Vendi score (VS). This raises the practical question of which one to use: LCR or VS. Here we address this question theoretically and numerically, using 53 large and well-known imaging and tabular datasets. We find that LCR and VS values can differ by orders of magnitude and are complementary, except in limiting cases. We show that both LCR and VS results depend on how similarities are scaled, and introduce the notion of ``half-distance'' to parameterize this dependence. We prove the VS provides an upper bound on LCR for all non-negative values of the Rényi-Hill order parameter, as well as for negative values in the special case that the similarity matrix is full rank. We conclude that VS is preferable only when a dataset's elements can be usefully interpreted as linear combinations of a more fundamental set of ``ur-elements'' or when the system that the dataset describes has a quantum-mechanical character. In the broader case where one simply wishes to capture the rich information encoded by elements' similarities and differences as well as their frequencies, we propose that LCR should be favored; nevertheless, for certain half-distances the two methods can complement each other.

cs.IT↗

arXiv2026

BanditLP: Large-Scale Stochastic Optimization for Personalized Recommendations

We present BanditLP, a scalable multi-stakeholder contextual bandit framework that unifies neural Thompson Sampling for learning objective-specific outcomes with a large-scale linear program for constrained action selection at serving time. The methodology is application-agnostic, compatible with arbitrary neural architectures, and deployable at web scale, with an LP solver capable of handling billions of variables. Experiments on public benchmarks and synthetic data show consistent gains over strong baselines. We apply this approach in LinkedIn's email marketing system and demonstrate business win, illustrating the value of integrated exploration and constrained optimization in production.

cs.LG↗

arXiv2026

A Synoptic Review of High-Frequency Oscillations as a Biomarker in Neurodegenerative Disease

High Frequency Oscillations (HFOs), rapid bursts of brain activity above 80 Hz, have emerged as a highly specific biomarker for epileptogenic tissue. Recent evidence suggests that HFOs are also present in Alzheimer's Disease (AD), reflecting underlying network hyperexcitability and offering a promising, noninvasive tool for early diagnosis and disease tracking. This synoptic review provides a comprehensive analysis of publicly available electroencephalography (EEG) datasets relevant to HFO research in neurodegenerative disorders. We conducted a bibliometric analysis of 1,222 articles, revealing a significant and growing research interest in HFOs, particularly within the last ten years. We then systematically profile and compare key public datasets, evaluating their participant cohorts, data acquisition parameters, and accessibility, with a specific focus on their technical suitability for HFO analysis. Our comparative synthesis highlights critical methodological heterogeneity across datasets, particularly in sampling frequency and recording paradigms, which poses challenges for cross-study validation, but also offers opportunities for robustness testing. By consolidating disparate information, clarifying nomenclature, and providing a detailed methodological framework, this review serves as a guide for researchers aiming to leverage public data to advance the role of HFOs as a cross-disease biomarker for AD and related conditions.

eess.SP↗

arXiv2026

$\textit{sentropy}$: A Python Package for Revealing Hidden Differences in Complex Datasets

Machine-learning datasets are typically characterized by measuring their size and class balance. However, there exists a richer and potentially more useful set of measures, termed S-entropy (similarity-sensitive entropy), that incorporate elements' frequencies and between-element similarities. Although these have been available in the R and Julia programming languages for other applications, they have not been as readily available in Python, which is widely used for machine learning, and are not easily applied to machine-learning-sized datasets without special coding considerations. To address these issues, we developed $\textit{sentropy}$, a Python package that calculates S-entropy and is tailored to large datasets. $\textit{sentropy}$ can calculate any of the frequency-sensitive measures of Hill's D-number framework and their similarity-sensitive counterparts. $\textit{sentropy}$ also outputs measures that compare datasets. We first briefly review S-entropy, illustrating how it incorporates elements' frequencies and elements' pairwise similarities. We then describe $\textit{sentropy}$'s key features and usage. We end with several examples - immunomics, metagenomics, computational pathology, and medical imaging - illustrating $\textit{sentropy}$'s applicability across a range of dataset types and fields.

q-bio.QM↗

arXiv2025

Beyond Basic A/B testing: Improving Statistical Efficiency for Business Growth

The standard A/B testing approaches are mostly based on t-test in large scale industry applications. These standard approaches however suffers from low statistical power in business settings, due to nature of small sample-size or non-Gaussian distribution or return-on-investment (ROI) consideration. In this paper, we (i) show the statistical efficiency of using estimating equation and U statistics, which can address these issues separately; and (ii) propose a novel doubly robust generalized U that allows flexible definition of treatment effect, and can handles small samples, distribution robustness, ROI and confounding consideration in one framework. We provide theoretical results on asymptotics and efficiency bounds, together with insights on the efficiency gain from theoretical analysis. We further conduct comprehensive simulation studies, apply the methods to multiple real A/B tests at LinkedIn, and share results and learnings that are broadly useful.

stat.ME↗

arXiv2025

TUS-REC2024: A Challenge to Reconstruct 3D Freehand Ultrasound Without External Tracker

Trackerless freehand ultrasound reconstruction aims to reconstruct 3D volumes from sequences of 2D ultrasound images without relying on external tracking systems. By eliminating the need for optical or electromagnetic trackers, this approach offers a low-cost, portable, and widely deployable alternative to more expensive volumetric ultrasound imaging systems, particularly valuable in resource-constrained clinical settings. However, predicting long-distance transformations and handling complex probe trajectories remain challenging. The TUS-REC2024 Challenge establishes the first benchmark for trackerless 3D freehand ultrasound reconstruction by providing a large publicly available dataset, along with a baseline model and a rigorous evaluation framework. By the submission deadline, the Challenge had attracted 43 registered teams, of which 6 teams submitted 21 valid dockerized solutions. The submitted methods span a wide range of approaches, including the state space model, the recurrent model, the registration-driven volume refinement, the attention mechanism, and the physics-informed model. This paper provides a comprehensive background introduction and literature review in the field, presents an overview of the challenge design and dataset, and offers a comparative analysis of submitted methods across multiple evaluation metrics. These analyses highlight both the progress and the current limitations of state-of-the-art approaches in this domain and provide insights for future research directions. All data and code are publicly available to facilitate ongoing development and reproducibility. As a live and evolving benchmark, it is designed to be continuously iterated and improved. The Challenge was held at MICCAI 2024 and is organised again at MICCAI 2025, reflecting its sustained commitment to advancing this field.

eess.IV↗

arXiv2025

GEM-T: Generative Tabular Data via Fitting Moments

Tabular data dominates data science but poses challenges for generative models, especially when the data is limited or sensitive. We present a novel approach to generating synthetic tabular data based on the principle of maximum entropy -- MaxEnt -- called GEM-T, for ``generative entropy maximization for tables.'' GEM-T directly captures nth-order interactions -- pairwise, third-order, etc. -- among columns of training data. In extensive testing, GEM-T matches or exceeds deep neural network approaches previously regarded as state-of-the-art in 23 of 34 publicly available datasets representing diverse subject domains (68\%). Notably, GEM-T involves orders-of-magnitude fewer trainable parameters, demonstrating that much of the information in real-world data resides in low-dimensional, potentially human-interpretable correlations, provided that the input data is appropriately transformed first. Furthermore, MaxEnt better handles heterogeneous data types (continuous vs. discrete vs. categorical), lack of local structure, and other features of tabular data. GEM-T represents a promising direction for light-weight high-performance generative models for structured data.

cs.LG↗