arXiv subjects

Gang Zhou

Publications and source records attributed to Gang Zhou.

At least 19 recordsLinked to original sources

arXiv2026

Retrieval Grounding Latent Reasoning for Dense Retrieval

Reasoning-intensive retrieval requires text representations to capture not only semantic similarity, but also the reasoning needed to determine relevance under a given retrieval instruction. Existing reasoning-enhanced embedding models improve retrieval by incorporating reasoning information into dense representations, yet their supervision is typically dominated by the final retrieval objective. As a result, latent reasoning trajectories may learn shortcut reasoning patterns that preserve retrieval performance without producing meaningful incremental retrieval gains. We propose Retrieval Grounding Latent Reasoning (RGLT), a latent reasoning framework for dense retrieval that explicitly connects intermediate latent transitions with retrieval improvements. RGLT performs non-autoregressive reasoning in hidden space through an instruction-conditioned latent reasoning trajectory constructed from silent tokens. It combines process-supervised explicit-to-implicit distillation with retrieval-grounded supervision, using stage-wise CoT reconstruction to shape intermediate latent states and retrieval-effect credit to optimize incremental retrieval gains across the latent reasoning trajectories. Experiments on reasoning-intensive retrieval benchmarks show that RGLT consistently outperforms strong baselines while preserving efficient embedding inference.

cs.AI↗

arXiv2026

Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAs

Vision-language models and vision-language action models endow the robot with unprecedented capabilities. However, the input of video and high-resolution images yields a massive number of visual tokens, leading to extremely high inference latency and severely hindering the robot's real-time control. To break through this computational bottleneck, we propose ST-Merge, a plug-and-play, training-free framework that efficiently fuses redundant tokens directly during the visual encoding phase. By explicitly constructing 3D spatiotemporal coordinates, it employs a multi-queue parallel matching and weighted aggregation mechanism to achieve efficient and geometrically consistent fusion of redundant tokens across frames. In addition, we introduce a post-merge positional correction mechanism that effectively eliminates spatial deviation caused by merging by dynamically re-evaluating the rotational position code of the weighted centroid of the vision token, thereby ensuring the high-precision spatial awareness required for dexterous operation. In the Video Question Answering task on the mainstream VLM, Qwen2.5-VL, ST-Merge achieves a 2$\times$ inference speedup with only a tiny 1\% loss in precision. When deployed on the $\pi_{0.5}$ VLA policy, ST-Merge achieves an 8.3$\times$ speedup at 1024 $\times$ 1024 resolution and matches the baseline success rate at this high-resolution setting. At lower resolutions, it introduces a small drop in accuracy.

cs.CV↗

arXiv2026

Sparse Autoencoders for Interpretable Emotion Control in Text-to-Speech

Integrating large language models (LLMs) into text-to-speech (TTS) systems has improved speech expressiveness, yet interpretable emotional control remains challenging. Existing approaches primarily rely on external conditioning or global activation steering, offering limited insight into the internal representations underlying emotional control. In this work, we analyze emotion-related variation in the semantic hidden states of LLM-based TTS models using sparse autoencoders (SAEs) to identify sparse latent features. Our analysis shows that emotional variation is distributed across multiple sparse latent features, while intervening on a small subset enables interpretable emotion control. Building on this observation, we introduce a feature-level intervention framework for bidirectional emotion induction and suppression without modifying backbone parameters. We further show that distinct latent features are associated with specific acoustic attributes (e.g., pitch), suggesting that emotional expression arises from coordinated latent contributions rather than a single global shift. Empirically, steering these sparse latent features achieves comparable or superior emotion induction and suppression performance relative to global steering and existing TTS baselines.

cs.CL↗

arXiv2025

FM-FoG: A Real-Time Foundation Model-based Wearable System for Freezing-of-Gait Mitigation

Freezing-of-Gait (FoG) affects over 50% of mid-to-late stage Parkinson's disease (PD) patients, significantly impairing patients' mobility independence and reducing quality of life. FoG is characterized by sudden episodes where walking cannot start or is interrupted, occurring exclusively during standing or walking, and never while sitting or lying down. Current FoG detection systems require extensive patient-specific training data and lack generalization, limiting clinical deployment. To address these issues, we introduce FM-FoG, a real-time foundation model-based wearable system achieving FoG detection in unseen patients without patient-specific training. Our approach combines self-supervised pretraining on diverse Inertial Measurement Unit (IMU) datasets with sensor context integration. Since FoG occurs only during ambulatory activities, a lightweight CNN-LSTM activity classifier selectively activates the foundation model only during walking or standing, avoiding unnecessary computation. Evaluated on the VCU FoG-IMU dataset with 23 PD patients, FM-FoG achieves a 98.5% F1-score when tested on previously unseen patients, substantially outperforming competitive baseline methods. Deployed on a Google Pixel 8a smartphone, the system extends battery life by up to 72% while maintaining sub-20ms intervention latency. The results indicate that our FM-FoG can enable practical, energy-efficient healthcare applications that generalize across patients without individual training requirements.

cs.LG↗

arXiv2025

All-in-One Medical Image Restoration with Latent Diffusion-Enhanced Vector-Quantized Codebook Prior

All-in-one medical image restoration (MedIR) aims to address multiple MedIR tasks using a unified model, concurrently recovering various high-quality (HQ) medical images (e.g., MRI, CT, and PET) from low-quality (LQ) counterparts. However, all-in-one MedIR presents significant challenges due to the heterogeneity across different tasks. Each task involves distinct degradations, leading to diverse information losses in LQ images. Existing methods struggle to handle these diverse information losses associated with different tasks. To address these challenges, we propose a latent diffusion-enhanced vector-quantized codebook prior and develop \textbf{DiffCode}, a novel framework leveraging this prior for all-in-one MedIR. Specifically, to compensate for diverse information losses associated with different tasks, DiffCode constructs a task-adaptive codebook bank to integrate task-specific HQ prior features across tasks, capturing a comprehensive prior. Furthermore, to enhance prior retrieval from the codebook bank, DiffCode introduces a latent diffusion strategy that utilizes the diffusion model's powerful mapping capabilities to iteratively refine the latent feature distribution, estimating more accurate HQ prior features during restoration. With the help of the task-adaptive codebook bank and latent diffusion strategy, DiffCode achieves superior performance in both quantitative metrics and visual quality across three MedIR tasks: MRI super-resolution, CT denoising, and PET synthesis.

cs.CV↗

arXiv2025

Evaluating an LLM-Powered Chatbot for Cognitive Restructuring: Insights from Mental Health Professionals

Recent advancements in large language models (LLMs) promise to expand mental health interventions by emulating therapeutic techniques, potentially easing barriers to care. Yet there is a lack of real-world empirical evidence evaluating the strengths and limitations of LLM-enabled psychotherapy interventions. In this work, we evaluate an LLM-powered chatbot, designed via prompt engineering to deliver cognitive restructuring (CR), with 19 users. Mental health professionals then examined the resulting conversation logs to uncover potential benefits and pitfalls. Our findings indicate that an LLM-based CR approach has the capability to adhere to core CR protocols, prompt Socratic questioning, and provide empathetic validation. However, issues of power imbalances, advice-giving, misunderstood cues, and excessive positivity reveal deeper challenges, including the potential to erode therapeutic rapport and ethical concerns. We also discuss design implications for leveraging LLMs in psychotherapy and underscore the importance of expert oversight to mitigate these concerns, which are critical steps toward safer, more effective AI-assisted interventions.

cs.HC↗

arXiv2024

IMUVIE: Pickup Timeline Action Localization via Motion Movies

Falls among seniors due to difficulties with tasks such as picking up objects pose significant health and safety risks, impacting quality of life and independence. Reliable, accessible assessment tools are critical for early intervention but often require costly clinic-based equipment and trained personnel, limiting their use in daily life. Existing wearable-based pickup measurement solutions address some needs but face limitations in generalizability. We present IMUVIE, a wearable system that uses motion movies and a machine-learning model to automatically detect and measure pickup events, providing a practical solution for frequent monitoring. IMUVIE's design principles-data normalization, occlusion handling, and streamlined visuals-enhance model performance and are adaptable to tasks beyond pickup classification. In rigorous leave one subject out cross validation evaluations, IMUVIE achieves exceptional window level localization accuracy of 91-92% for pickup action classification on 256,291 motion movie frame candidates while maintaining an event level recall of 97% when evaluated on 129 pickup events. IMUVIE has strong generalization and performs well on unseen subjects. In an interview survey, IMUVIE demonstrated strong user interest and trust, with ease of use identified as the most critical factor for adoption. IMUVIE offers a practical, at-home solution for fall risk assessment, facilitating early detection of movement deterioration, and supporting safer, independent living for seniors.

cs.LG↗

arXiv2024

GPT-based Textile Pilling Classification Using 3D Point Cloud Data

Textile pilling assessment is critical for textile quality control. We collect thousands of 3D point cloud images in the actual test environment of textiles and organize and label them as TextileNet8 dataset. To the best of our knowledge, it is the first publicly available eight-categories 3D point cloud dataset in the field of textile pilling assessment. Based on PointGPT, the GPT-like big model of point cloud analysis, we incorporate the global features of the input point cloud extracted from the non-parametric network into it, thus proposing the PointGPT+NN model. Using TextileNet8 as a benchmark, the experimental results show that the proposed PointGPT+NN model achieves an overall accuracy (OA) of 91.8% and a mean per-class accuracy (mAcc) of 92.2%. Test results on other publicly available datasets also validate the competitive performance of the proposed PointGPT+NN model. The proposed TextileNet8 dataset will be publicly available.

cs.CV↗

arXiv2024

Enhancing Adversarial Robustness of Vision-Language Models through Low-Rank Adaptation

Vision-Language Models (VLMs) play a crucial role in the advancement of Artificial General Intelligence (AGI). As AGI rapidly evolves, addressing security concerns has emerged as one of the most significant challenges for VLMs. In this paper, we present extensive experiments that expose the vulnerabilities of conventional adaptation methods for VLMs, highlighting significant security risks. Moreover, as VLMs grow in size, the application of traditional adversarial adaptation techniques incurs substantial computational costs. To address these issues, we propose a parameter-efficient adversarial adaptation method called \textbf{\textit{AdvLoRA}} based on Low-Rank Adaptation. We investigate and reveal the inherent low-rank properties involved in adversarial adaptation for VLMs. Different from LoRA, we enhance the efficiency and robustness of adversarial adaptation by introducing a novel reparameterization method that leverages parameter clustering and alignment. Additionally, we propose an adaptive parameter update strategy to further bolster robustness. These innovations enable our AdvLoRA to mitigate issues related to model security and resource wastage. Extensive experiments confirm the effectiveness and efficiency of AdvLoRA.

cs.CV↗

arXiv2024

NetBench: A Large-Scale and Comprehensive Network Traffic Benchmark Dataset for Foundation Models

In computer networking, network traffic refers to the amount of data transmitted in the form of packets between internetworked computers or Cyber-Physical Systems. Monitoring and analyzing network traffic is crucial for ensuring the performance, security, and reliability of a network. However, a significant challenge in network traffic analysis is to process diverse data packets including both ciphertext and plaintext. While many methods have been adopted to analyze network traffic, they often rely on different datasets for performance evaluation. This inconsistency results in substantial manual data processing efforts and unfair comparisons. Moreover, some data processing methods may cause data leakage due to improper separation of training and testing data. To address these issues, we introduce the NetBench, a large-scale and comprehensive benchmark dataset for assessing machine learning models, especially foundation models, in both network traffic classification and generation tasks. NetBench is built upon seven publicly available datasets and encompasses a broad spectrum of 20 tasks, including 15 classification tasks and 5 generation tasks. Furthermore, we evaluate eight State-Of-The-Art (SOTA) classification models (including two foundation models) and two generative models using our benchmark. The results show that foundation models significantly outperform the traditional deep learning methods in traffic classification. We believe NetBench will facilitate fair comparisons among various approaches and advance the development of foundation models for network traffic. Our benchmark is available at https://github.com/WM-JayLab/NetBench.

cs.NI↗

arXiv2024

Lens: A Knowledge-Guided Foundation Model for Network Traffic

Network traffic refers to the amount of data being sent and received over the Internet or any system that connects computers. Analyzing network traffic is vital for security and management, yet remains challenging due to the heterogeneity of plain-text packet headers and encrypted payloads. To capture the latent semantics of traffic, recent studies have adopted Transformer-based pretraining techniques to learn network representations from massive traffic data. However, these methods pre-train on data-driven tasks but overlook network knowledge, such as masking partial digits of the indivisible network port numbers for prediction, thereby limiting semantic understanding. In addition, they struggle to extend classification to new classes during fine-tuning due to the distribution shift. Motivated by these limitations, we propose \Lens, a unified knowledge-guided foundation model for both network traffic classification and generation. In pretraining, we propose a Knowledge-Guided Mask Span Prediction method with textual context for learning knowledge-enriched representations. For extending to new classes in finetuning, we reframe the traffic classification as a closed-ended generation task and introduce context-aware finetuning to adapt to the distribution shift. Evaluation results across various benchmark datasets demonstrate that the proposed Lens~achieves superior performance on both classification and generation tasks. For traffic classification, Lens~outperforms competitive baselines substantially on 8 out of 12 tasks with an average accuracy of \textbf{96.33\%} and extends to novel classes with significantly better performance. For traffic generation, Lens~generates better high-fidelity network traffic for network simulation, gaining up to \textbf{30.46\%} and \textbf{33.3\%} better accuracy and F1 in fuzzing tests. We will open-source the code upon publication.

cs.LG↗

arXiv2023

SANDFORMER: CNN and Transformer under Gated Fusion for Sand Dust Image Restoration

Although Convolutional Neural Networks (CNN) have made good progress in image restoration, the intrinsic equivalence and locality of convolutions still constrain further improvements in image quality. Recent vision transformer and self-attention have achieved promising results on various computer vision tasks. However, directly utilizing Transformer for image restoration is a challenging task. In this paper, we introduce an effective hybrid architecture for sand image restoration tasks, which leverages local features from CNN and long-range dependencies captured by transformer to improve the results further. We propose an efficient hybrid structure for sand dust image restoration to solve the feature inconsistency issue between Transformer and CNN. The framework complements each representation by modulating features from the CNN-based and Transformer-based branches rather than simply adding or concatenating features. Experiments demonstrate that SandFormer achieves significant performance improvements in synthetic and real dust scenes compared to previous sand image restoration methods.

cs.CV↗

arXiv2022

DeepAuditor: Distributed Online Intrusion Detection System for IoT devices via Power Side-channel Auditing

As the number of IoT devices has increased rapidly, IoT botnets have exploited the vulnerabilities of IoT devices. However, it is still challenging to detect the initial intrusion on IoT devices prior to massive attacks. Recent studies have utilized power side-channel information to identify this intrusion behavior on IoT devices but still lack accurate models in real-time for ubiquitous botnet detection. We proposed the first online intrusion detection system called DeepAuditor for IoT devices via power auditing. To develop the real-time system, we proposed a lightweight power auditing device called Power Auditor. We also designed a distributed CNN classifier for online inference in a laboratory setting. In order to protect data leakage and reduce networking redundancy, we then proposed a privacy-preserved inference protocol via Packed Homomorphic Encryption and a sliding window protocol in our system. The classification accuracy and processing time were measured, and the proposed classifier outperformed a baseline classifier, especially against unseen patterns. We also demonstrated that the distributed CNN design is secure against any distributed components. Overall, the measurements were shown to the feasibility of our real-time distributed system for intrusion detection on IoT devices.

cs.CR↗

arXiv2021

GRIM: A General, Real-Time Deep Learning Inference Framework for Mobile Devices based on Fine-Grained Structured Weight Sparsity

It is appealing but challenging to achieve real-time deep neural network (DNN) inference on mobile devices because even the powerful modern mobile devices are considered as ``resource-constrained'' when executing large-scale DNNs. It necessitates the sparse model inference via weight pruning, i.e., DNN weight sparsity, and it is desirable to design a new DNN weight sparsity scheme that can facilitate real-time inference on mobile devices while preserving a high sparse model accuracy. This paper designs a novel mobile inference acceleration framework GRIM that is General to both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) and that achieves Real-time execution and high accuracy, leveraging fine-grained structured sparse model Inference and compiler optimizations for Mobiles. We start by proposing a new fine-grained structured sparsity scheme through the Block-based Column-Row (BCR) pruning. Based on this new fine-grained structured sparsity, our GRIM framework consists of two parts: (a) the compiler optimization and code generation for real-time mobile inference; and (b) the BCR pruning optimizations for determining pruning hyperparameters and performing weight pruning. We compare GRIM with Alibaba MNN, TVM, TensorFlow-Lite, a sparse implementation based on CSR, PatDNN, and ESE (a representative FPGA inference acceleration framework for RNNs), and achieve up to 14.08x speedup.

cs.LG↗

arXiv2021

On the non-degenerate and degenerate generic singularities formed by mean curvature flow

In this paper we study a neighborhood of generic singularities formed by mean curvature flow (MCF). We limit our consideration to the singularities modelled on $\mathbb{S}^3\times\mathbb{R}$ because, compared to the cases $\mathbb{S}^k\times \mathbb{R}^{l}$ with $l\geq 2$, the present case has the fewest possibilities to be considered. For various possibilities, we provide a detailed description for a small, but fixed, neighborhood of singularity, and prove that a small neighborhood of the singularity is mean convex, and the singularity is isolated. For the remaining possibilities, we conjecture that an entire neighborhood of the singularity becomes singular at the time of blowup, and present evidences to support this conjecture. A key technique is that, when looking for a dominating direction for the rescaled MCF, we need a normal form transformation, as a result, the rescaled MCF is parametrized over some chosen curved cylinder, instead over a standard straight one. This is a long paper. The introduction is carefully written to present the key steps and ideas.

math.DG↗

arXiv2019

Gate controlled quantum interference: direct observation of anti-resonances in single molecule charge transport

Quantum interference can profoundly affect charge transport in single molecules, but experiments can usually measure only the conductance at the Fermi energy. Because in general the most pronounced features of the quantum interference are not located at the Fermi energy, it is highly desirable to probe charge transport in a broader energy range. Here by the method of electrochemical gating, we measure the conductance and map the transmission functions of single molecules at and around the Fermi energy, and study signatures associated with constructive and destructive interference. With the electrochemical gate control, we tune the quantum interference between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO), and directly observe anti-resonance, a distinct feature of destructive interference. By tuning the molecule in and out of anti-resonance, we achieve continuous control of the conductance over 2 orders of magnitude with a subthreshold swing of ~17 mV/dec, features relevant to high-speed and low-power electronics.

cond-mat.mes-hall↗

arXiv2016

HMOG: New Behavioral Biometric Features for Continuous Authentication of Smartphone Users

We introduce Hand Movement, Orientation, and Grasp (HMOG), a set of behavioral features to continuously authenticate smartphone users. HMOG features unobtrusively capture subtle micro-movement and orientation dynamics resulting from how a user grasps, holds, and taps on the smartphone. We evaluated authentication and biometric key generation (BKG) performance of HMOG features on data collected from 100 subjects typing on a virtual keyboard. Data was collected under two conditions: sitting and walking. We achieved authentication EERs as low as 7.16% (walking) and 10.05% (sitting) when we combined HMOG, tap, and keystroke features. We performed experiments to investigate why HMOG features perform well during walking. Our results suggest that this is due to the ability of HMOG features to capture distinctive body movements caused by walking, in addition to the hand-movement dynamics from taps. With BKG, we achieved EERs of 15.1% using HMOG combined with taps. In comparison, BKG using tap, key hold, and swipe features had EERs between 25.7% and 34.2%. We also analyzed the energy consumption of HMOG feature extraction and computation. Our analysis shows that HMOG features extracted at 16Hz sensor sampling rate incurred a minor overhead of 7.9% without sacrificing authentication accuracy. Two points distinguish our work from current literature: 1) we present the results of a comprehensive evaluation of three types of features (HMOG, keystroke, and tap) and their combinations under the same experimental conditions, and 2) we analyze the features from three perspectives (authentication, BKG, and energy consumption on smartphones).

cs.CR↗

arXiv2012

Dirac Fermion in Strongly-Bound Graphene Systems

It is highly desirable to integrate graphene into existing semiconductor technology, where the combined system is thermodynamically stable yet maintain a Dirac cone at the Fermi level. Firstprinciples calculations reveal that a certain transition metal (TM) intercalated graphene/SiC(0001), such as the strongly-bound graphene/intercalated-Mn/SiC, could be such a system. Different from free-standing graphene, the hybridization between graphene and Mn/SiC leads to the formation of a dispersive Dirac cone of primarily TM d characters. The corresponding Dirac spectrum is still isotropic, and the transport behavior is nearly identical to that of free-standing graphene for a bias as large as 0.6 V, except that the Fermi velocity is half that of graphene. A simple model Hamiltonian is developed to qualitatively account for the physics of the transfer of the Dirac cone from a dispersive system (e.g., graphene) to an originally non-dispersive system (e.g., TM).

cond-mat.mes-hall↗