arXiv subjects

Jeffery Wu

Publications and source records attributed to Jeffery Wu.

6 recordsLinked to original sources

arXiv2026

LPH-VTON: Resolving the Structure-Texture Dilemma of Virtual Try-On via Latent Process Handover

Virtual Try-On (VTON) aims to synthesize photorealistic images of garments precisely aligned with a person's body and pose. Current diffusion-based methods, however, face a fundamental trade-off between structural integrity and textural fidelity. In this paper, we formalize this challenge as a consequence of complementary inductive biases inherent in prevailing architectures: models heavily reliant on spatial constraints naturally favor geometric alignment but often suppress textures, whereas models dominated by unconstrained generative priors excel at vibrant detail rendering but are prone to structural drift. Based on this diagnosis, we propose LPH-VTON, a new synergistic framework that resolves this tension within a single, continuous denoising process. LPH-VTON strategically decomposes the generation, leveraging a structure-biased model to establish a geometrically consistent latent scaffold in the early stages, before handing over control to a texture-biased model for high-fidelity detail rendering. Extensive experiments validate our approach. Our model achieves a superior Pareto-optimal balance, establishing new benchmarks in perceptual faithfulness while maintaining highly competitive structural alignment across the standard dataset VITON-HD, proving the efficacy of temporal architectural decoupling.

cs.CV↗

arXiv2026

ClawGuard: Out-of-Band Detection of LLM Agent Workflow Hijacking via EM Side Channel

Autonomous LLM agents face a critical security risk known as workflow hijacking, where attackers subtly alter tool and skill invocations. Existing defenses rely on host-internal telemetry (such as audit logs), which can be forged if the host OS is compromised. To solve this, we introduce ClawGuard, a passive, out-of-band monitor that audits LLM-agent workflows using electromagnetic (EM) emanations. Because distinct agent skills create unique hardware usage patterns (computation, DRAM, network blocking), they emit measurable, macroscopic EM envelopes. External software-defined radios (SDRs) capture these physical signals. Using a drift-aware pipeline with 320-dimensional features, ClawGuard converts RF streams into physical evidence. Evaluated on a 7.82TB RF corpus, ClawGuard achieved an AUC of 0.9945, detecting attacks with a 100% true-positive rate and a 1.16% false-positive rate. This proves passive EM sensing is a practical, forge-resistant physical check against compromised host software.

cs.CR↗

arXiv2026

A measure for genuine tripartite entanglement

We introduce a single real-valued functional $I(\vec{n}_1,\vec{n}_2)$, built from four three-qubit correlation expectation values, that turns the Greenberger--Horne--Zeilinger (GHZ) algebraic paradox into a \emph{quantitative} witness of genuine tripartite entanglement. We prove that for every three-qubit state $\rho$ and every pair of measurement directions $|I(\vec{n}_1,\vec{n}_2;\rho)|\le 2$, with the bound saturated if and only if $\vec{n}_1\perp \vec{n}_2$ and $\rho$ is locally unitarily equivalent to the GHZ state. We obtain a closed-form expression for $I(\hat{x},\hat{y})$ on the five-parameter Ac\'in canonical family of three-qubit pure states; it depends only on the product $\lambda_0\lambda_4$ and is maximised when $\lambda_0=\lambda_4=1/\sqrt{2}$. For the W state we show that $I(\hat{x},\hat{y})=0$ and that $\max_{\vec{n}_1, \vec{n}_2} | I_{W} |=35/27 \approx 1.296$, strictly below the GHZ value. Maximising the underlying correlation structure over \emph{independent} local orthonormal frames on the three parties yields a manifestly local-unitary (LU) invariant quantity $\mathcal{E}_{GHZ}(\rho)\in[0,1]$ that equals one if and only if $\rho$ is LU equivalent to the GHZ state, takes the value $35/54\approx 0.648$ on the W state, and is bounded by $1/2$ on all biseparable and fully separable states; it is therefore a device-independent indicator of GHZ-type genuine tripartite correlation. We carefully delimit which properties are proven and which (notably global convexity and the resulting genuine-multipartite-entanglement witness threshold) are established numerically and remain open analytically. We also outline a generalisation of $I$ to three-qudit systems built from the Heisenberg--Weyl operators, recovering the standard qubit construction when $d=2$.

quant-ph↗

arXiv2026

How to quantify direct correlations between variables

A crucial question throughout statistics is whether an observed correlation between two variables is a direct correlation or only an indirect one mediated by a confounder. We organize the existing nonlinear measures of direct correlation into two families, each with a systematic construction: (i) removing the direct correlation from the joint distribution and quantifying the resulting distributional shift, and (ii) intervening on one variable via do-calculus and quantifying the response of the other. For every Kullback-Leibler-based measure in either family we propose a Jensen-Shannon-based regularized analogue; the regularized measures take values in $[0,1]$, satisfy the metric property, and are free of the singularities of the Kullback-Leibler divergence. We analyze the achievable upper bound of each regularized measure under the observed marginals, and derive the maximal value each measure can attain when only the alphabet sizes of the variables are fixed; the maxima admit closed forms built on a single binary-entropy function. The measures are compared on a decision-making model and on three public datasets (Titanic survival, UCI Adult income, and the 1973 Berkeley graduate admissions), with bootstrap confidence intervals for every reported value.

stat.ME↗

arXiv2025

One Walk is All You Need: Data-Efficient 3D RF Scene Reconstruction with Human Movements

Reconstructing 3D Radiance Field (RF) scenes through opaque obstacles is a long-standing goal, yet it is fundamentally constrained by a laborious data acquisition process requiring thousands of static measurements, which treats human motion as noise to be filtered. This work introduces a new paradigm with a core objective: to perform fast, data-efficient, and high-fidelity RF reconstruction of occluded 3D static scenes, using only a single, brief human walk. We argue that this unstructured motion is not noise, but is in fact an information-rich signal available for reconstruction. To achieve this, we design a factorization framework based on composite 3D Gaussian Splatting (3DGS) that learns to model the dynamic effects of human motion from the persistent static scene geometry within a raw RF stream. Trained on just a single 60-second casual walk, our model reconstructs the full static scene with a Structural Similarity Index (SSIM) of 0.96, remarkably outperforming heavily-sampled state-of-the-art (SOTA) by 12%. By transforming the human movements into its valuable signals, our method eliminates the data acquisition bottleneck and paves the way for on-the-fly 3D RF mapping of unseen environments.

cs.NI↗

arXiv2025

Enhancing Self-Driving Segmentation in Adverse Weather Conditions: A Dual Uncertainty-Aware Training Approach to SAM Optimization

Recent advances in vision foundation models, such as the Segment Anything Model (SAM) and its successor SAM2, have achieved state-of-the-art performance on general image segmentation benchmarks. However, these models struggle in adverse weather conditions where visual ambiguity is high, largely due to their lack of uncertainty quantification. Inspired by progress in medical imaging, where uncertainty-aware training has improved reliability in ambiguous cases, we investigate two approaches to enhance segmentation robustness for autonomous driving. First, we introduce a multi-step finetuning procedure for SAM2 that incorporates uncertainty metrics directly into the loss function, improving overall scene recognition. Second, we adapt the Uncertainty-Aware Adapter (UAT), originally designed for medical image segmentation, to driving contexts. We evaluate both methods on CamVid, BDD100K, and GTA driving datasets. Experiments show that UAT-SAM outperforms standard SAM in extreme weather, while SAM2 with uncertainty-aware loss achieves improved performance across diverse driving scenes. These findings underscore the value of explicit uncertainty modeling for safety-critical autonomous driving in challenging environments.

cs.CV↗