Source author record

Ziyu Huang

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Artificial Intelligence math.PR Information Theory Machine Learning math.IT math.OC

Catalog footprint

Research overview

4works

6topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2026arXiv

A Class of Degenerate Mean Field Games, Associated FBSDEs and Master Equations

In this paper, we study a class of degenerate mean field games (MFGs) with state-distribution dependent and unbounded functional diffusion coefficients. With a probabilistic method, we study the well-posedness of the forward-backward stochastic differential equations (FBSDEs) associated with the MFG and arising from the maximum principle, and estimate the corresponding Jacobian and Hessian flows. We further establish the classical regularity of the value functional $V$ ; in particular, we show that when the cost function is $C^3$ in the spatial and control variables and $C^2$ in the distribution argument, then the value functional is $C^1$ in time and $C^2$ in the spatial and distribution variables. As a consequence, the value functional $V$ is the unique classical solution of the degenerate MFG master equation.

preprint2026arXiv

Channel Knowledge Map Construction via Guided Flow Matching

The efficient construction of accurate channel knowledge maps (CKMs) is crucial for unleashing the full potential of environment-aware wireless networks, yet it remains a difficult ill-posed problem due to the sparsity of available location-specific channel knowledge data. Although diffusion-based methods such as denoising diffusion probabilistic models (DDPMs) have been exploited for CKM construction, they rely on iterative stochastic sampling, rendering them too slow for real-time wireless applications. To bridge the gap between high fidelity and efficient CKM construction, this letter introduces a novel framework based on linear transport guided flow matching (LT-GFM). Deviating from the noise-removal paradigm of diffusion models, our approach models the CKM generation process as a deterministic ordinary differential equation (ODE) that follows linear optimal transport paths, thereby drastically reducing the number of required inference steps. We propose a unified architecture that is applicable to not only the conventional channel gain map (CGM) construction, but also the more challenging spatial correlation map (SCM) construction. To achieve physics-informed CKM constructions, we integrate environmental semantics (e.g., building masks) for edge recovery and enforce Hermitian symmetry for property of the SCM. Simulation results verify that LT-GFM achieves superior distributional fidelity with significantly lower Fréchet Inception Distance (FID) and accelerates inference speed by a factor of 25 compared to DDPMs.

preprint2026arXiv

RPO: Fine-Tuning Visual Generative Models via Rich Vision-Language Preferences

Traditional preference tuning methods for LLMs/Visual Generative Models often rely solely on reward model labeling, which can be opaque, offer limited insights into the rationale behind preferences, and are prone to issues such as reward hacking or overfitting. We introduce Rich Preference Optimization (RPO), a novel pipeline that leverages rich feedback signals from Vision Language Models (VLMs) to improve the curation of preference pairs for fine-tuning visual generative models like text-to-image diffusion models. Our approach begins with prompting VLMs to generate detailed critiques of synthesized images, from which we further prompt VLMs to extract reliable and actionable image editing instructions. By implementing these instructions, we create refined images, resulting in synthetic, informative preference pairs that serve as enhanced tuning datasets. We demonstrate the effectiveness of our pipeline and the resulting datasets in fine-tuning state-of-the-art diffusion models.

preprint2022arXiv

State-Density Flows of Non-Degenerate Density-Dependent Mean Field SDEs and Associated PDEs

In this paper, we study a combined system of a Fokker-Planck (FP) equation for $m^{t,μ}$ with initial $(t,μ)\in[0,T]\times L^2(\mathbb{R}^d)$ , and a stochastic differential equation for $X^{t,x,μ}$ with initial $(t,x)\in[0,T]\times \mathbb{R}^d$ , whose coefficients depend on the solution of FP equation. We develop a combined probabilistic and analytical method to explore the regularity of the functional $V(t,x,μ)=\mathbb{E}[Φ(X^{t,x,μ}_T,m^{t,μ}(T,\cdot))]$ . Our main result states that, under a non-degenerate condition and appropriate regularity assumptions on the coefficients, the function $V$ is the unique classical solution of a nonlocal partial differential equation of mean-field type. The proof depends heavily on the differential properties of the flow $μ\mapsto (m^{t,μ}, X^{t,x,μ})$ over $μ\in L^2(\mathbb{R}^d)$ . We also give an example to illustrate the role of our main result. Finally, we give a discussion on the $L^1$ choice case in the initial $μ$ .

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Ziyu Huang

Research overview

Is this your work?

4 publications

A Class of Degenerate Mean Field Games, Associated FBSDEs and Master Equations

Channel Knowledge Map Construction via Guided Flow Matching

RPO: Fine-Tuning Visual Generative Models via Rich Vision-Language Preferences

State-Density Flows of Non-Degenerate Density-Dependent Mean Field SDEs and Associated PDEs

See the researcher in context

Building this map preview