arXiv subjects

Kaiwen Guo

Publications and source records attributed to Kaiwen Guo.

At least 19 recordsLinked to original sources

arXiv2026

Extremals and Thresholds for Critical Singular Anisotropic Moser-Trudinger Inequalities

Let $N\ge2$, $q>1$, and $0<β 0,\qquad 0 0$. For $q_-\le q a_c$ if $a_c<\infty$. The argument does not determine attainment at $a=a_c$. Radial blow-up analysis gives a concentration bound in terms of a nonlinear Green function. The first nonzero Taylor term determines the strict comparison, an exact Euclidean reduction shows that the threshold is independent of $F$, a separate supplement gives an alternative proof for $q=N$ using the classical Green function and compact Taylor terms.

math.AP↗

arXiv2026

StrucPhysVideo: Learning Physical Dynamics from Structured Captions and Robot Actions

Modeling physical dynamics, including how objects move, interact, and change state, is central to video world models for embodied AI. We present StrucPhysVideo, a family of video world models that bridges physics-focused data curation with language- and action-conditioned prediction of scene evolution. Our data pipeline combines motion-aware video segmentation, quality and content filtering, and physical relevance verification with structured annotations of objects, materials, and temporally localized interactions. By disentangling camera motion from object behavior and explicitly describing contact, deformation, and state transitions, the pipeline provides supervision grounded in observable physical events. Building on these data, we introduce StrucPhysVideo-TI2V, a sparse Mixture-of-Experts (MoE) text-image-to-video model trained with a curriculum that progressively emphasizes physical dynamics while retaining general-domain video data. StrucPhysVideo-TI2V achieves state-of-the-art performance on Physics-IQ Verified, scoring 45.5% and outperforming Cosmos3-Super-Image2Video by 2.8 percentage points. Caption ablations across backbones further demonstrate the effectiveness of physics-focused supervision. We further extend StrucPhysVideo-TI2V to StrucPhysVideo-IA2V, an interactive image-action-to-video world model that predicts visual outcomes from robot end-effector commands. Action conditioning, causal autoregressive generation, and few-step distillation enable incremental robot rollouts with only four denoising steps. Together, StrucPhysVideo advances physical dynamics modeling from image- and language-conditioned video prediction toward action-driven interaction.

cs.CV↗

arXiv2026

CCL-Bench 1.0: A Trace-Based Benchmark for LLM Infrastructure

Evaluative claims about LLM infrastructure -- ``workload X is fastest on hardware Y with software Z'' -- depend on a complex configuration space spanning hardware accelerators, interconnect bandwidth, software frameworks, parallelism plans, and communication libraries. Current infrastructure evaluation benchmarks publish a small set of end-to-end numbers that do not explain why one configuration outperforms another. We present CCL-Bench, a trace-based benchmark that addresses the limitations of existing benchmarks by recording reusable evidence for every ML workload. Each contributed data point in CCL-Bench packages an execution trace, a YAML workload card, and the launch scripts. We have developed a community-extensible toolkit to compute fine-grained compute, memory, and communication efficiency metrics from this evidence. Using CCL-Bench, we surface three claims that summary-statistic benchmarks cannot support: (i) higher compute-communication overlap can coincide with longer training step time and reveal inefficient parallelization choices, (ii) doubling TPU interconnect bandwidth yields a much higher end-to-end improvement in step time than doubling GPU interconnect bandwidth on small and medium workloads, and (iii) the best-tuned configuration on one training framework can run up to 3$\times$ slower than the best-tuned configuration on a peer framework on identical hardware.

cs.DC↗

arXiv2026

SUAD: Solid-Channel Ultrasound Injection Attack and Defense to Voice Assistants

As a versatile AI application, voice assistants (VAs) have become increasingly popular, but are vulnerable to security threats. Attackers have proposed various inaudible attacks, but are limited by cost, distance, or LoS. Therefore, we propose \name~Attack, a long-range, cross-barrier, and interference-free inaudible voice attack via solid channels. We begin by thoroughly analyzing the dispersion effect in solid channels, revealing its unique impact on signal propagation. To avoid distortions in voice commands, we design a modular command generation model that parameterizes attack distance, victim audio, and medium dispersion features to adapt to variations in the solid-channel state. Additionally, we propose SUAD Defense, a universal defense that uses ultrasonic perturbation signals to block inaudible voice attacks (IVAs) without impacting normal speech. Since the attack can occur at arbitrary frequencies and times, we propose a training method that randomizes both time and frequency to generate perturbation signals that break ultrasonic commands. Notably, the perturbation signal is modulated to an inaudible frequency without affecting the functionality of voice commands for VAs. Experiments on six smartphones have shown that SUAD Attack achieves activation success rates above 89.8% and SUAD Defense blocks IVAs with success rates exceeding 98%.

cs.CR↗

arXiv2026

MindGuard: Intrinsic Decision Inspection for Securing LLM Agents Against Metadata Poisoning

The Model Context Protocol (MCP) is increasingly adopted to standardize the interaction between LLM agents and external tools. However, this trend introduces a new threat: Tool Poisoning Attacks (TPA), where tool metadata is poisoned to induce the agent to perform unauthorized operations. Existing defenses that primarily focus on behavior-level analysis are fundamentally ineffective against TPA, as poisoned tools need not be executed, leaving no behavioral trace to monitor. Thus, we propose MindGuard, a decision-level guardrail for LLM agents, providing provenance tracking of call decisions, policy-agnostic detection, and poisoning source attribution against TPA. While fully explaining LLM decision remains challenging, our empirical findings uncover a strong correlation between LLM attention mechanisms and tool invocation decisions. Therefore, we choose attention as an empirical signal for decision tracking and formalize this as the Decision Dependence Graph (DDG), which models the LLM's reasoning process as a weighted, directed graph where vertices represent logical concepts and edges quantify the attention-based dependencies. We further design robust DDG construction and graph-based anomaly analysis mechanisms that efficiently detect and attribute TPA attacks. Extensive experiments on real-world datasets demonstrate that MindGuard achieves 94\%-99\% average precision in detecting poisoned invocations, 95\%-100\% attribution accuracy, with processing times under one second and no additional token cost. Moreover, DDG can be viewed as an adaptation of the classical Program Dependence Graph (PDG), providing a solid foundation for applying traditional security policies at the decision level.

cs.CR↗

arXiv2025

Existence of extremal functions and Wulff symmetry for anisotropic Trudinger-Moser inequalities

In this paper, we investigate the extremal functions for anisotropic Trudinger-Moser inequalities. Our method uses convex symmetrization, the continuity of the supremum function, together with the relation between the supremums of the subcritical and the critical anisotropic Trudinger-Moser inequality, we give some results of existence and symmetry about the extremal functions for several different types of anisotropic Trudinger-Moser inequalities.

math.FA↗

arXiv2025

AttenTrack: Mobile User Attention Awareness Based on Context and External Distractions

In the mobile internet era, managing limited attention amid information overload is crucial for enhancing collaboration and information delivery. However, current attention-aware systems often depend on wearables or personalized data, limiting their scalability and cross-context adaptability. Inspired by psychological theories, we attempt to treat mobile notifications as naturally occurring external distractions and infer users' attention states based on their response behaviors and contextual information. Our goal is to build an attention-aware model that does not rely on personalized historical data or complex subjective input, while ensuring strong cold-start capability and cross-context adaptability. To this end, We design a field study framework integrating subjective and objective data, closely aligned with real-world external distractions (i.e., mobile notifications). Through field studies, we construct a fine-grained and interpretable dataset centered on the relationship among current context - external distractions - subjective attention. Through our field studies, we conduct an in-depth analysis of the relationships among users' response behaviors, response motivations, contextual information, and attention states. Building on our findings, we propose AttenTrack, a lightweight, privacy-friendly attention awareness model with strong cold-start capability. The model relies solely on non-privacy-sensitive objective data available on mobile devices, and can be applied to a variety of attention management tasks. In addition, we will publicly release the constructed dataset to support future research and advance the field of mobile attention awareness.

cs.HC↗

arXiv2025

Ascending the Infinite Ladder: Benchmarking Spatial Deformation Reasoning in Vision-Language Models

Humans naturally possess the spatial reasoning ability to form and manipulate images and structures of objects in space. There is an increasing effort to endow Vision-Language Models (VLMs) with similar spatial reasoning capabilities. However, it remains unclear whether these models truly understand and manipulate spatial objects or not. To address this question, we propose a new evaluation framework aimed at assessing the performance of VLMs in spatial deformation reasoning tasks. Specifically, we construct a benchmark for spatial deformation reasoning from 2D to 3D. Leveraging our data engine, we can generate unlimited evaluation problem pairs with infinite steps, without any data leakage. We explore whether the model can effectively perform spatial deformation reasoning from two directions: forward reasoning (given the operations, find the final state) and reverse reasoning (given the final state, determine the operations). We adopt a ladder competition format, using the number of deformation steps as the level classification criterion, with the goal of exploring the boundaries of the model's deformation reasoning capabilities. Interestingly, the benchmarking results reveal that almost no model demonstrates plausible spatial deformation reasoning abilities. Furthermore, even after applying targeted training and mainstream reasoning enhancement methods, the models are still unable to perform well on 3D spatial deformation reasoning.

cs.CV↗

arXiv2025

SECNEURON: Reliable and Flexible Abuse Control in Local LLMs via Hybrid Neuron Encryption

Large language models (LLMs) with diverse capabilities are increasingly being deployed in local environments, presenting significant security and controllability challenges. These locally deployed LLMs operate outside the direct control of developers, rendering them more susceptible to abuse. Existing mitigation techniques mainly designed for cloud-based LLM services are frequently circumvented or ineffective in deployer-controlled environments. We propose SECNEURON, the first framework that seamlessly embeds classic access control within the intrinsic capabilities of LLMs, achieving reliable, cost-effective, flexible, and certified abuse control for local deployed LLMs. SECNEURON employs neuron-level encryption and selective decryption to dynamically control the task-specific capabilities of LLMs, limiting unauthorized task abuse without compromising others. We first design a task-specific neuron extraction mechanism to decouple logically related neurons and construct a layered policy tree for handling coupled neurons. We then introduce a flexible and efficient hybrid encryption framework for millions of neurons in LLMs. Finally, we developed a distribution-based decrypted neuron detection mechanism on ciphertext to ensure the effectiveness of partially decrypted LLMs. We proved that SECNEURON satisfies IND-CPA Security and Collusion Resistance Security under the Task Controllability Principle. Experiments on various task settings show that SECNEURON limits unauthorized task accuracy to below 25% while keeping authorized accuracy loss with 2%. Using an unauthorized Code task example, the accuracy of abuse-related malicious code generation was reduced from 59% to 15%. SECNEURON also mitigates unauthorized data leakage, reducing PII extraction rates to below 5% and membership inference to random guesses.

cs.CR↗

arXiv2024

Singular Trudinger--Moser inequality involving $L^{p}$ norm in bounded domain

In this paper, we use the method of blow-up analysis and capacity estimate to derive the singular Trudinger--Moser inequality involving $N$-Finsler--Laplacian and $L^{p}$ norm, precisely, for any $p>1$, $0\leqγ<γ_{1}:= \inf\limits_{u\in W^{1, N}_{0}(Ω)\backslash \{0\}}\frac{\int_ΩF^{N}(\nabla u)dx}{\| u\|_p^N}$ and $0\leqβ<N$, we have \begin{align} \sup_{u\in W_{0}^{1,N}(Ω),\;\int_ΩF^{N}(\nabla u)dx-γ\| u\|_p^N\leq1}\int_Ω\frac{e^{λ_{N}(1-\fracβ{N})\lvert u\rvert^{\frac{N}{N-1}}}}{F^{o}(x)^β}\;\mathrm{d}x<+\infty\notag, \end{align} where $λ_{N}=N^{\frac{N}{N-1}} κ_{N}^{\frac{1}{N-1}}$ and $κ_{N}$ is the volume of a unit Wulff ball in $\mathbb{R}^N$, moreover, extremal functions for the inequality are also obtained. When $F=\lvert\cdot\rvert$ and $p=N$, we can obtain the singular version of Tintarev type inequality by the obove inequality, namely, for any $0\leqα<α_{1}(Ω):=\inf\limits_{u\in W^{1, N}_{0}(Ω)\backslash \{0\}}\frac{\int_Ω|\nabla u|^Ndx}{\| u\|_N^N}$ and $0\leqβ<N$, it holds $$ \sup_{u\in W_{0}^{1,N}(Ω),\;\int_Ω\lvert\nabla u\rvert^{N}\;\mathrm{d}x-α\|u\|_{N}^{N}\leq1}\int_Ω\frac{e^{α_{N}(1-\fracβ{N})\lvert u\rvert^{\frac{N}{N-1}}}}{\lvert x\rvert^β}\;\mathrm{d}x<+\infty, $$ where $α_{N}:=N^{\frac{N}{N-1}}ω_{N}^{\frac{1}{N-1}}$ and $ ω_{N}$ is the volume of unit ball in $\mathbb{R}^{N}$. Our results extend many well-known Trudinger--Moser type inequalities to more general setting.

math.AP↗

arXiv2024

SGSM: A Foundation-model-like Semi-generalist Sensing Model

The significance of intelligent sensing systems is growing in the realm of smart services. These systems extract relevant signal features and generate informative representations for particular tasks. However, building the feature extraction component for such systems requires extensive domain-specific expertise or data. The exceptionally rapid development of foundation models is likely to usher in newfound abilities in such intelligent sensing. We propose a new scheme for sensing model, which we refer to as semi-generalist sensing model (SGSM). SGSM is able to semiautomatically solve various tasks using relatively less task-specific labeled data compared to traditional systems. Built through the analysis of the common theoretical model, SGSM can depict different modalities, such as the acoustic and Wi-Fi signal. Experimental results on such two heterogeneous sensors illustrate that SGSM functions across a wide range of scenarios, thereby establishing its broad applicability. In some cases, SGSM even achieves better performance than sensor-specific specialized solutions. Wi-Fi evaluations indicate a 20\% accuracy improvement when applying SGSM to an existing sensing model.

eess.SP↗

arXiv2024

URHand: Universal Relightable Hands

Existing photorealistic relightable hand models require extensive identity-specific observations in different views, poses, and illuminations, and face challenges in generalizing to natural illuminations and novel identities. To bridge this gap, we present URHand, the first universal relightable hand model that generalizes across viewpoints, poses, illuminations, and identities. Our model allows few-shot personalization using images captured with a mobile phone, and is ready to be photorealistically rendered under novel illuminations. To simplify the personalization process while retaining photorealism, we build a powerful universal relightable prior based on neural relighting from multi-view images of hands captured in a light stage with hundreds of identities. The key challenge is scaling the cross-identity training while maintaining personalized fidelity and sharp details without compromising generalization under natural illuminations. To this end, we propose a spatially varying linear lighting model as the neural renderer that takes physics-inspired shading as input feature. By removing non-linear activations and bias, our specifically designed lighting model explicitly keeps the linearity of light transport. This enables single-stage training from light-stage data while generalizing to real-time rendering under arbitrary continuous illuminations across diverse identities. In addition, we introduce the joint learning of a physically based model and our neural relighting model, which further improves fidelity and generalization. Extensive experiments show that our approach achieves superior performance over existing methods in terms of both quality and generalizability. We also demonstrate quick personalization of URHand from a short phone scan of an unseen identity.

cs.CV↗

arXiv2023

Drivable Avatar Clothing: Faithful Full-Body Telepresence with Dynamic Clothing Driven by Sparse RGB-D Input

Clothing is an important part of human appearance but challenging to model in photorealistic avatars. In this work we present avatars with dynamically moving loose clothing that can be faithfully driven by sparse RGB-D inputs as well as body and face motion. We propose a Neural Iterative Closest Point (N-ICP) algorithm that can efficiently track the coarse garment shape given sparse depth input. Given the coarse tracking results, the input RGB-D images are then remapped to texel-aligned features, which are fed into the drivable avatar models to faithfully reconstruct appearance details. We evaluate our method against recent image-driven synthesis baselines, and conduct a comprehensive analysis of the N-ICP algorithm. We demonstrate that our method can generalize to a novel testing environment, while preserving the ability to produce high-fidelity and faithful clothing dynamics and appearance.

cs.GR↗

arXiv2023

Sharp anisotropic singular Trudinger-Moser inequalities in the entire space

In this paper, we investigate sharp singular Trudinger-Moser inequalities involving the anisotropic Dirichlet norm $\left(\int_ΩF^{N}(\nabla u)\;\mathrm{d}x\right)^{\frac{1}{N}}$ in the Sobolev-type space $D^{N,q}(\mathbb{R}^{N})$, $q\geq 1$, here $F:\mathbb{R}^{N}\rightarrow[0,+\infty)$ is a convex function of class $C^{2}(\mathbb{R}^{N}\setminus\{0\})$, which is even and positively homogeneous of degree 1, its polar $F^{0}$ represents a Finsler metric on $\mathbb{R}^{N}$. Combing with the connection between convex symmetrization and Schwarz symmetrization, we will establish anisotropic singular Trudinger-Moser inequalities and discuss their sharpness under several different situations, including the case $\|F(\nabla u)\|_{N}\leq 1$, the case $\|F(\nabla u)\|_{N}^{a}+\|u\|_{q}^{b}\leq 1$, and whether they are associated with exact growth.

math.FA↗

arXiv2023

Grouped Knowledge Distillation for Deep Face Recognition

Compared with the feature-based distillation methods, logits distillation can liberalize the requirements of consistent feature dimension between teacher and student networks, while the performance is deemed inferior in face recognition. One major challenge is that the light-weight student network has difficulty fitting the target logits due to its low model capacity, which is attributed to the significant number of identities in face recognition. Therefore, we seek to probe the target logits to extract the primary knowledge related to face identity, and discard the others, to make the distillation more achievable for the student network. Specifically, there is a tail group with near-zero values in the prediction, containing minor knowledge for distillation. To provide a clear perspective of its impact, we first partition the logits into two groups, i.e., Primary Group and Secondary Group, according to the cumulative probability of the softened prediction. Then, we reorganize the Knowledge Distillation (KD) loss of grouped logits into three parts, i.e., Primary-KD, Secondary-KD, and Binary-KD. Primary-KD refers to distilling the primary knowledge from the teacher, Secondary-KD aims to refine minor knowledge but increases the difficulty of distillation, and Binary-KD ensures the consistency of knowledge distribution between teacher and student. We experimentally found that (1) Primary-KD and Binary-KD are indispensable for KD, and (2) Secondary-KD is the culprit restricting KD at the bottleneck. Therefore, we propose a Grouped Knowledge Distillation (GKD) that retains the Primary-KD and Binary-KD but omits Secondary-KD in the ultimate KD loss calculation. Extensive experimental results on popular face recognition benchmarks demonstrate the superiority of proposed GKD over state-of-the-art methods.

cs.CV↗

arXiv2023

Size Effect and Scaling in Quasi-static and Fatigue Fracture of Graphene Polymer Nanocomposites

This work investigated how the structure size affects the quasi-static and fatigue behaviors of graphene polymer nanocomposites, a topic that has been often overlooked. The results showed that both quasi-static and fatigue failure of these materials scale nonlinearly with the structure size due to the presence of a significant Fracture Process Zone (FPZ) ahead of the crack tip induced by graphene nanomodification. Such a complicated size effect and scaling in either quasi-static or fatigue scenario cannot be described by the Linear Elastic Fracture Mechanics (LEFM), but can be well captured by the Size Effect Law (SEL) which considers the FPZ. Thanks to the SEL, the enhanced quasi-static and fatigue fracture properties were properly characterized and shown to be independent of the structure size. In addition, the differences on the morphological and mechanical behaviors between quasi-static fracture and fatigue fracture were also identified and clarified in this work. The experimental data and analytical analyses reported in this paper are important to deeply understand the mechanics of polymer-based nanocomposite materials and even other quasi-brittle materials (e.g., fiber-reinforced polymers or its hybrid with nanoparticles, etc.), and further advance the development of computational models capable of capturing size-dependent fracture of materials in various loading conditions.

physics.app-ph↗

arXiv2022

Drivable Volumetric Avatars using Texel-Aligned Features

Photorealistic telepresence requires both high-fidelity body modeling and faithful driving to enable dynamically synthesized appearance that is indistinguishable from reality. In this work, we propose an end-to-end framework that addresses two core challenges in modeling and driving full-body avatars of real people. One challenge is driving an avatar while staying faithful to details and dynamics that cannot be captured by a global low-dimensional parameterization such as body pose. Our approach supports driving of clothed avatars with wrinkles and motion that a real driving performer exhibits beyond the training corpus. Unlike existing global state representations or non-parametric screen-space approaches, we introduce texel-aligned features -- a localised representation which can leverage both the structural prior of a skeleton-based parametric model and observed sparse image signals at the same time. Another challenge is modeling a temporally coherent clothed avatar, which typically requires precise surface tracking. To circumvent this, we propose a novel volumetric avatar representation by extending mixtures of volumetric primitives to articulated objects. By explicitly incorporating articulation, our approach naturally generalizes to unseen poses. We also introduce a localized viewpoint conditioning, which leads to a large improvement in generalization of view-dependent appearance. The proposed volumetric representation does not require high-quality mesh tracking as a prerequisite and brings significant quality improvements compared to mesh-based counterparts. In our experiments, we carefully examine our design choices and demonstrate the efficacy of our approach, outperforming the state-of-the-art methods on challenging driving scenarios.

cs.CV↗

arXiv2022

Geometry-aware Single-image Full-body Human Relighting

Single-image human relighting aims to relight a target human under new lighting conditions by decomposing the input image into albedo, shape and lighting. Although plausible relighting results can be achieved, previous methods suffer from both the entanglement between albedo and lighting and the lack of hard shadows, which significantly decrease the realism. To tackle these two problems, we propose a geometry-aware single-image human relighting framework that leverages single-image geometry reconstruction for joint deployment of traditional graphics rendering and neural rendering techniques. For the de-lighting, we explore the shortcomings of UNet architecture and propose a modified HRNet, achieving better disentanglement between albedo and lighting. For the relighting, we introduce a ray tracing-based per-pixel lighting representation that explicitly models high-frequency shadows and propose a learning-based shading refinement module to restore realistic shadows (including hard cast shadows) from the ray-traced shading maps. Our framework is able to generate photo-realistic high-frequency shadows such as cast shadows under challenging lighting conditions. Extensive experiments demonstrate that our proposed method outperforms previous methods on both synthetic and real images.

cs.CV↗