arXiv subjects

Zhen Liu

Publications and source records attributed to Zhen Liu.

At least 55 records · Page 3Linked to original sources

arXiv2026

CamFlow+: Hybrid Motion Bases for 2D Camera Motion Estimation with Stabilization Applications

Estimating 2D camera motion is fundamental to computer vision and computational photography. Existing homography-based methods work well for planar scenes or pure rotation, but struggle with camera translation, depth variation, and local parallax; local homography and mesh-based models improve flexibility but still rely on piecewise planar assumptions. We introduce CamFlow+, a hybrid-basis framework that represents 2D camera motion directly in dense-flow space. CamFlow+ combines homography-derived physical bases, stochastic bases sampled from homography flows, and depth-translational bases derived from depth and camera intrinsics, relaxing the single-plane constraint while preserving camera-motion regularity. A depth-aware smoothness term further regularizes translation-induced parallax in continuous-depth regions while preserving motion changes near depth boundaries. We evaluate CamFlow+ on GHOF-Cam, a camera-motion benchmark that masks out dynamic objects and ill-posed occlusion regions in an optical-flow benchmark to isolate camera-induced motion. Experiments show that CamFlow+ improves sparse and dense camera-motion estimation. In digital video stabilization, CamFlow+ also improves global and local stability, achieving the best top-1 preference rate in a blind user study. Code and datasets will be available on the project page: https://lhaippp.github.io/CamFlow+.

cs.CV↗

arXiv2026

Drifting Preference Optimization for One-Step Generative Models

One-step text-to-image generators are attractive for deployment because they generate an image with a single forward pass, but preference finetuning them remains difficult: standard alignment methods often rely on policy likelihoods, denoising trajectories, differentiable reward gradients, or test-time optimization. We propose Drifting Preference Optimization (DrPO), an online preference-finetuning method for deterministic one-step generators. For each prompt, DrPO samples candidates from the current generator, ranks them with a target reward, and uses high- and low-scoring samples to synthesize a feature-space update direction. The update is a non-parametric dipole preference field plus a reference drift estimated from the frozen base generator, and is optimized through a detached feature-space regression target. The target reward is used only for ranking, so DrPO can train with large, black-box, or non-differentiable rewards while inference remains a single generator call. We evaluate DrPO on SD-Turbo and SDXL-Turbo with multiple target rewards and benchmarks, including HPSv3 and GenEval. DrPO improves alignment over reward-gradient-free one-step preference baselines and reduces HPSv3 training computation by $3.51\times$ under the matched effective-batch setting by removing reward-model backpropagation. Initial offline experiments suggest that sample-based gradient synthesis can also be used beyond online reward ranking.

cs.LG↗

arXiv2026

Towards Anatomically Plausible Human Image Generation via Synthetic Localized Preferences

Large-scale text-to-image foundation models have achieved remarkable visual realism, yet generating human images with correct anatomical structures remains challenging. Existing approaches enforce anatomical constraints through part-specific modules or localized loss weighting during supervised fine-tuning on high-quality human photos, but such datasets are limited and often provide ambiguous optimization signals due to confounding factors such as lighting, pose, and background. Preference-based alignment offers an alternative, but standard Direct Preference Optimization (DPO) treats all pixels equally and therefore fails to exploit the localized nature of anatomical artifacts. To address this, we propose the framework of Alignment via Synthetic Anatomical Preference (ASAP), which constructs controlled preference pairs through a localized degradation mechanism applied to high-fidelity human images. This mechanism performs a controlled experiment on images by introducing explicit anatomical errors in targeted regions while preserving the remaining content. With this mechanism, we create the Human Anatomical Preference (HAP) dataset with over 10K curated pairs for effective anatomical alignment of text-to-image human image generative models. To better leverage the locality of these controlled preference pairs, we introduce a localized and margin-bounded variant of DPO that prioritizes optimization in targeted anatomical regions while enforcing a finite preference margin to prevent over-optimization and preserve global semantics. We further introduce HAF-Bench, a benchmark for systematic evaluation of anatomical fidelity. Extensive experiments demonstrate that ASAP consistently reduces anatomical errors across multiple foundation models while maintaining overall image quality.

cs.CV↗

arXiv2026

APS: Bias-Controlled Adaptive Prototype Simulation for Population-Scale LLM Agents

LLM-agent simulation offers a flexible computational tool for studying population response trajectories that depend on scenario events, memory, demographics, and evolving social context. However, full multi-round simulation scales linearly with both population size and horizon, requiring every agent to query the LLM at every round. We propose Adaptive Prototype Simulation (APS), a framework that reframes scalable LLM-based simulation as a recurrent oracle-allocation problem. APS retains the designated LLM as the online transition oracle while querying adaptive core prototypes, selected singleton-tail agents, and shadow-audit agents. Prototype responses induce local response surfaces for nearby agents, reducing online LLM calls without replacing the underlying transition model. To control approximation bias, shadow-audit residual correction estimates propagation residuals for aggregate correction and future budget allocation, while tail-protected singleton routing directly queries selected isolated, heterogeneous, or high-curvature regions that are vulnerable to smoothing. Theoretically, we treat APS as an estimator for full-scale high-precision individual social simulation and decompose its errors into prototype-coverage error, shadow-audit residual-correction error, local-propagation bias, and temporal context mismatch. Under the reported protocols, APS gives lower reference-aligned distributional discrepancy than scale-oriented and same-budget baselines while reducing online LLM calls, with ablations and compact robustness checks diagnosing the main bias-control mechanisms. In a 10M-agent, multi-round public-opinion simulation, APS achieves a 381.1-fold reduction over full simulation, with reference-aligned final-round JSD of 0.094 against the corresponding full-LLM reference.

cs.MA↗

arXiv2026

A New Source of Millicharged Particles: Secondary Showers in the LHC Forward Absorber

Millicharged particles (mCPs) are a well-motivated target for far-forward searches at the Large Hadron Collider. We identify and quantify a significant new source of these particles: secondary production in hadronic and electromagnetic showers initiated by energetic neutral particles striking the TAXN absorber. By combining Monte Carlo simulations with \texttt{Geant4}-based modeling, we show that these secondary cascades yield a substantial mCP flux that complements the primary production from the interaction point. For the proposed FORMOSA detector, this contribution can enhance the expected signal yield by approximately $50\%$ for $m_χ\lesssim 0.1~\textrm{GeV}$. Our results demonstrate that secondary production in downstream infrastructure is an essential ingredient for realistic sensitivity projections and new-physics searches at the High-Luminosity LHC. The simulated secondary spectra are made publicly available to facilitate future forward physics studies.

hep-ph↗

arXiv2026

ZeroIDIR: Zero-Reference Illumination Degradation Image Restoration with Perturbed Consistency Diffusion Models

In this paper, we propose a zero-reference diffusion-based framework, named ZeroIDIR, for illumination degradation image restoration, which decouples the restoration process into adaptive illumination correction and diffusion-based reconstruction while being trained solely on low-quality degraded images. Specifically, we design an adaptive gamma correction module that performs spatially varying exposure correction to generate illumination-corrected only representations to mitigate exposure bias and serve as reliable inputs for subsequent diffusion processes, where a histogram-guided illumination correction loss is introduced to regularize the corrected illumination distribution toward that of natural scenes. Subsequently, the illumination-corrected image is treated as an intermediate noisy state for the proposed perturbed consistency diffusion model to reconstruct details and suppress noise. Moreover, a perturbed diffusion consistency loss is proposed to constrain the forward diffusion trajectory of the final restored image to remain consistent with the perturbed state, thus improving restoration fidelity and stability in the absence of supervision. Extensive experiments on publicly available benchmarks show that the proposed method outperforms state-of-the-art unsupervised competitors and is comparable to supervised methods while being more generalizable to various scenes. Code is available at https://github.com/JianghaiSCU/ZeroIDIR.

cs.CV↗

arXiv2026

Electroweak Restoration: SMEFT and HEFT

Colliders continue to push our understanding of electroweak (EW) interactions to ever higher energies. At high energies, many observables in the broken EW theory are expected to approach the unbroken theory. This is the electroweak restoration regime, where longitudinal gauge boson production corresponds to Goldstone boson production. As such, in this paper we investigate electroweak restoration in the context of linear and non-linear realizations of the EW symmetry in longitudinal di-boson production: $f\bar{f}'\rightarrow V_LV'_L$ and $f\bar{f}'\rightarrow V_Lh$, where $V_L,V'_L$ are longitudinal gauge bosons, $h$ is the Higgs boson, and $f,f'$ are SM fermions. For the linear beyond the SM (BSM) theory, we use the Standard Model Effective Theory (SMEFT), and for the non-linear BSM theory, we use Higgs Effective Field Theory (HEFT). We give a general discussion of these amplitudes and cross sections in the SM, SMEFT, and HEFT. Using the Goldstone boson equivalence theorem, we derive a set of ratios among high energy $V_LV'_L$ and $V_Lh$ amplitudes that are expected to approach one in the SM and SMEFT. Through our explicit calculations, we show that these ratios do indeed approach one in the SM and dimension-6 SMEFT, but not necessarily HEFT. Beyond the amplitudes, both theoretically and experimentally, we identify the cross section ratio of $W^\pm_LZ_L$ and $W^\pm_L h$ as a particularly promising observable to probe EW restoration and distinguish HEFT and SMEFT. Using current LHC measurements as well as projections for $W^\pm_L h$ measurements, we project HL-LHC sensitivities to probing the linear vs. non-linear realizations of the EW symmetry.

hep-ph↗

arXiv2026

ExpoCM: Exposure-Aware One-Step Generative Single-Image HDR Reconstruction

Single-image HDR reconstruction aims to recover high dynamic range radiance from a single low dynamic range (LDR) input, but remains highly ill-posed due to detail saturation in over-exposed regions and noise amplification in under-exposed areas. While recent diffusion-based approaches offer powerful generative priors, they often overlook the exposure-dependent nature of the degradation and incur substantial computational costs from iterative sampling. To address these challenges, we propose ExpoCM, a novel one-step generative HDR reconstruction framework that reformulates HDR reconstruction as a Probability Flow ODE (PF-ODE) and constructs exposure-aware consistency trajectories via exposure-dependent perturbations. Specifically, a soft exposure mask is first constructed to separate the LDR image into over-, under-, and well-exposed regions. Based on this partition, region-conditioned consistency trajectories are designed to hallucinate saturated details, suppress noise in dark regions, and preserve reliable structures within a single, distillation-free inference step. To further enhance perceptual quality, we introduce an Exposure-guided Luminance-Chromaticity Loss in the CIE~$\text{L}^*\text{a}^*\text{b}^*$ space, which assigns exposure-aware weights to luminance and chromaticity components, effectively mitigating brightness bias and color drift. Extensive experiments on the HDR-REAL, HDR-EYE, and AIM2025 benchmarks demonstrate that ExpoCM achieves state-of-the-art fidelity and perceptual accuracy, while enabling over 400$\times$ and 20$\times$ faster inference compared to DDPM (1000 steps) and DDIM (50 steps), respectively.

cs.CV↗

arXiv2026

Flipping Against All Odds: Reducing LLM Coin Flip Bias via Verbalized Rejection Sampling

Large language models (LLMs) can often accurately describe probability distributions using natural language, yet they still struggle to generate faithful samples from them. This mismatch limits their use in tasks requiring reliable stochasticity, such as Monte Carlo methods, agent-based simulations, and randomized decision-making. We investigate this gap between knowledge and sampling in the context of Bernoulli distributions. We introduce Verbalized Rejection Sampling (VRS), a natural-language adaptation of classical rejection sampling that prompts the LLM to reason about and accept or reject proposed samples. Despite relying on the same Bernoulli mechanism internally, VRS substantially reduces sampling bias across models. We provide theoretical analysis showing that, under mild assumptions, VRS improves over direct sampling, with gains attributable to both the algorithm and prompt design. More broadly, our results show how classical probabilistic tools can be verbalized and embedded into LLM workflows to improve reliability, without requiring access to model internals or heavy prompt engineering.

cs.LG↗

arXiv2026

The Essence of Balance for Self-Improving Agents in Vision-and-Language Navigation

In vision-and-language navigation (VLN), self-improvement from policy-induced experience, using only standard VLN action supervision, critically depends on balancing behavioral diversity and learning stability, which governs whether the agent can extract a reliable learning signal for improvement. Increasing behavioral diversity is necessary to expose alternative action hypotheses but can destabilize policy-induced learning signals, whereas overly conservative stability constraints suppress exploration and induce early commitment, making reliable self-improvement difficult. To address this challenge, we propose Stability-Diversity Balance (SDB), a plug-and-play mechanism for balanced self-improvement in VLN. SDB expands each decision step into multiple latent behavioral hypotheses by applying controlled shifts in the instruction-conditioned hidden states, and then performs reliability-aware soft evaluation and aggregation to retain diverse yet instruction-consistent alternatives during learning. An explicit regularizer further constrains hypothesis interactions, preventing excessive drift or premature collapse of hypothesis diversity and stabilizing self-improvement without discarding training signals. Experiments on R2R, SOON, and REVERIE show consistent improvements; for example, on REVERIE val-unseen, SDB improves SPL from 33.73 to 35.93 and OSR from 51.07 to 54.25.

cs.CV↗

arXiv2026

Instruction-as-State: Environment-Guided and State-Conditioned Semantic Understanding for Embodied Navigation

Vision-and-Language Navigation requires agents to follow natural-language instructions in visually changing environments. A central challenge is the dynamic entanglement between language and observations: the meaning of instruction shifts as the agent's field of view and spatial context evolve. However, many existing models encode the instruction as a static global representation, limiting their ability to adapt instruction meaning to the current visual context. We therefore model instruction understanding as an Instruction-as-State variable: a decision-relevant, token-level instruction state that evolves step by step conditioned on the agent's perceptual state, where the perceptual state denotes the observation-grounded navigation context at each step. To realize this principle, we introduce State-Entangled Environment-Guided Instruction Understanding (S-EGIU), a coarse-to-fine framework for state-conditioned segment activation and token-level semantic refinement. At the coarse level, S-EGIU activates the instruction segment whose semantics align with the current observation. At the fine level, it refines the activated segment through observation-guided token grounding and contextual modeling, sharpening its internal semantics under the current observation. Together, these stages maintain an instruction state that is continuously updated according to the agent's perceptual state during navigation. S-EGIU delivers strong performance on several key metrics, including a +2.68% SPL gain on REVERIE Test Unseen, and demonstrates consistent efficiency gains across multiple VLN benchmarks, underscoring the value of dynamic instruction--perception entanglement.

cs.CV↗

arXiv2026

CoPA: Benchmarking Personalized Question Answering with Data-Informed Cognitive Factors

While LLMs have demonstrated remarkable potential in Question Answering (QA), evaluating personalization remains a critical bottleneck. Existing paradigms predominantly rely on lexical-level similarity or manual heuristics, often lacking sufficient data-driven validation. We address this by mining Community-Individual Preference Divergence (CIPD), where individual choices override consensus, to distill six key personalization factors as evaluative dimensions. Accordingly, we introduce CoPA, a benchmark with 1,985 user profiles for fine-grained, factor-level assessment. By quantifying the alignment between model outputs and user-specific cognitive preferences inferred from interaction patterns, CoPA provides a more comprehensive and discriminative standard for evaluating personalized QA than generic metrics. The code is available at https://github.com/bjzgcai/CoPA.

cs.CL↗

arXiv2026

A Busy Higgs Signal

Higgs final states are prime targets in the search for physics beyond the Standard Model. In the conventional picture, $SU(2)$ symmetry together with the Goldstone Equivalence Theorem correlates Higgs and gauge-boson final states, implying comparable sensitivity in channels such as $hh$, $ZZ$, and $WW$ in searches for heavy resonances. In this work, we identify a mechanism to parametrically violate this expectation. We show that higher-order Higgs couplings can induce an electroweak-symmetry-breaking enhancement that selectively amplifies Higgs-rich final states, allowing them to become the leading discovery channels of new resonances. For scalar resonances, this can make di-Higgs the dominant bosonic signal. For resonance masses higher than a couple of TeV, it also opens resonant tri-Higgs and four-Higgs channels as well-motivated search targets. The same underlying mechanism extends to heavy fermionic and vector resonances, where it can similarly enhance channels such as $ht$, $Zh$, and $γh$. We present this framework in effective field theory, demonstrate possible UV completions, and discuss its implications for collider searches.

hep-ph↗

arXiv2026

Pressure-Robust Fortin-Soulie Elements of the Stokes Equation on Curved Domains

This paper presents a pressure-robust and element-wise divergence-free nonconforming finite element method for the Stokes problem on curved domains. The discrete element is constructed by mapping the Fortin-Soulie element from a reference triangle using an isoparametric mapping for the geometry and a Piola transform for the function space. The inf-sup condition and the error estimate with optimal convergence rate are proved. Pressure-robustness is obtained by replacing the discrete velocity test functions with the first-order Raviart-Thomas functions. Numerical examples are provided to validate the theoretical results.

math.NA↗

arXiv2026

Paths of length five with equal-degree endpoints

Addressing a question posed by Erdős and Hajnal, Chen and Ma proved that, for all $n \ge 600$, the complete bipartite graph $K_{n,n+1}$ is the unique graph on $2n+1$ vertices with at least $n^2+n$ edges that contains no two vertices of equal degree joined by a path of length three. In this paper, we extend this result and show that, for all $n \ge 11$, $K_{n,n+1}$ is the unique $(2n+1)$-vertex graph with at least $n^2+n$ edges that avoids two equal-degree vertices joined by a path of length five. This confirms the very next case of a general conjecture of Chen and Ma on paths of odd length with equal-degree endpoints.

math.CO↗

arXiv2026

Degree sequences realizing labelled $h$-factors

For a positive integer $ k $, let $ [k] = \{1, 2, \ldots, k\} $. Let $ h $ be a non-negative integer, and let $ n $ be a multiple of $ h + 1 $. Define $ H $ as the disjoint union of $ n/(h+1) $ cliques (each of size $ h + 1 $) with vertex sets $ V_1, \ldots, V_{n/(h+1)} $, where $ V_i = \{ v_j \mid j = (i-1)(h+1) + k, k \in [h+1] \} $ for $ i \in [n/(h+1)] $. A non-increasing integer sequence $ (d_1, \ldots, d_n) $ is $ H $-realizable if there exists a graph $ G $ with $ V(G) = V(H) = \{ v_i \mid i \in [n] \} $, $ d_G(v_i) = d_i $ for all $ i\in [n] $, and $ G $ contains $ H $ as a spanning subgraph. If $ h = 0 $, then a non-increasing integer sequence $ (d_1, \ldots, d_n) $ is $ H $-realizable if and only if there exists a graph $ G $ with degree sequence $ (d_1, d_2, \dots, d_n) $; Erdős and Gallai established a necessary and sufficient condition for this property. Recently, Briggs, McDonald, and Shan extended their result to the case $ h = 1 $. In this paper, we establish a necessary and sufficient condition for a sequence $ (d_1, d_2, \dots, d_n) $ to be $ H $-realizable for any non-negative integer $ h $, thereby confirming a conjecture due to Briggs, McDonald and Shan.

math.CO↗

arXiv2026

EVA: Aligning Video World Models with Executable Robot Actions via Inverse Dynamics Rewards

Video generative models are increasingly used as world models for robotics, where a model generates a future visual rollout conditioned on the current observation and task instruction, and an inverse dynamics model (IDM) converts the generated frames into executable robot actions. However, current video world models lack explicit executability constraints. As a result, visually coherent rollouts may still violate rigid-body and kinematic consistency, producing unstable or infeasible control commands when decoded by an IDM. We refer to this mismatch between visual generation and physically executable control as the executability gap. While this gap can be mitigated at inference time using techniques such as rejection sampling, such approaches are inefficient due to the high cost of video generation. In this paper, we leverage the executability gap as a training signal and introduce Executable Video Alignment (EVA), a reinforcement-learning post-training framework for aligning video world models. EVA trains an inverse dynamics model on real robot trajectories and repurposes it as a reward model that evaluates generated videos through the action sequences they induce, encouraging smooth motions measured by velocity, acceleration, and jerk while penalizing actions that violate embodiment constraints. Importantly, the reward remains informative even when generated videos contain severe visual artifacts, since such artifacts typically translate into unstable or out-of-bound actions. Experiments on the RoboTwin benchmark and a real bimanual robot show that EVA reduces embodiment-specific artifacts in generated rollouts and improves downstream task execution success.

cs.RO↗

arXiv2026

Improved Dark Photon Sensitivity from the Dark SRF Experiment

We report the refined dark-photon exclusion bound from Dark SRF's pathfinder run. Our new result is driven by improved theoretical modeling of frequency instability in high-quality resonant experiments. Our analysis leads to a constraint that is an order of magnitude stronger than previously reported (corresponding to a signal-to-noise ratio that is four orders of magnitude larger). This result represents the world-leading constraint on non-dark-matter dark photons over a wide range of masses below $6\,\rm μeV$ and translates to the best laboratory-based limit on the photon mass $m_γ<2.9\times 10^{-48}\,\rm g$.

hep-ex↗