arXiv subjects

Weiqi Zhang

Publications and source records attributed to Weiqi Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2025

MaterialRefGS: Reflective Gaussian Splatting with Multi-view Consistent Material Inference

Modeling reflections from 2D images is essential for photorealistic rendering and novel view synthesis. Recent approaches enhance Gaussian primitives with reflection-related material attributes to enable physically based rendering (PBR) with Gaussian Splatting. However, the material inference often lacks sufficient constraints, especially under limited environment modeling, resulting in illumination aliasing and reduced generalization. In this work, we revisit the problem from a multi-view perspective and show that multi-view consistent material inference with more physically-based environment modeling is key to learning accurate reflections with Gaussian Splatting. To this end, we enforce 2D Gaussians to produce multi-view consistent material maps during deferred shading. We also track photometric variations across views to identify highly reflective regions, which serve as strong priors for reflection strength terms. To handle indirect illumination caused by inter-object occlusions, we further introduce an environment modeling strategy through ray tracing with 2DGS, enabling photorealistic rendering of indirect radiance. Experiments on widely used benchmarks show that our method faithfully recovers both illumination and geometry, achieving state-of-the-art rendering quality in novel views synthesis.

cs.CV↗

arXiv2025

SynerGen: Contextualized Generative Recommender for Unified Search and Recommendation

The dominant retrieve-then-rank pipeline in large-scale recommender systems suffers from mis-calibration and engineering overhead due to its architectural split and differing optimization objectives. While recent generative sequence models have shown promise in unifying retrieval and ranking by auto-regressively generating ranked items, existing solutions typically address either personalized search or query-free recommendation, often exhibiting performance trade-offs when attempting to unify both. We introduce \textit{SynerGen}, a novel generative recommender model that bridges this critical gap by providing a single generative backbone for both personalized search and recommendation, while simultaneously excelling at retrieval and ranking tasks. Trained on behavioral sequences, our decoder-only Transformer leverages joint optimization with InfoNCE for retrieval and a hybrid pointwise-pairwise loss for ranking, allowing semantic signals from search to improve recommendation and vice versa. We also propose a novel time-aware rotary positional embedding to effectively incorporate time information into the attention mechanism. \textit{SynerGen} achieves significant improvements on widely adopted recommendation and search benchmarks compared to strong generative recommender and joint search and recommendation baselines. This work demonstrates the viability of a single generative foundation model for industrial-scale unified information access.

cs.CL↗

arXiv2025

Empowering Time Series Analysis with Foundation Models: A Comprehensive Survey

Time series data are ubiquitous across diverse real-world applications, making time series analysis critically important. Traditional approaches are largely task-specific, offering limited functionality and poor transferability. In recent years, foundation models have revolutionized NLP and CV with their remarkable cross-task transferability, zero-/few-shot learning capabilities, and multimodal integration capacity. This success has motivated increasing efforts to explore foundation models for addressing time series modeling challenges. Although some tutorials and surveys were published in the early stages of this field, the rapid pace of recent developments necessitates a more comprehensive and in-depth synthesis to cover the latest advances. Our survey aims to fill this gap by introducing a modality-aware, challenge-oriented perspective, which reveals how foundation models pre-trained on different modalities face distinct hurdles when adapted to time series tasks. Building on this perspective, we propose a taxonomy of existing works organized by pre-training modality (time series, language, and vision), analyze modality-specific challenges and categorize corresponding solutions, discussing their advantages and limitations. Beyond this, we review real-world applications to illustrate domain-specific advancements, provide open-source codes, and conclude with potential future research directions in this rapidly evolving field.

cs.LG↗

arXiv2025

MedualTime: A Dual-Adapter Language Model for Medical Time Series-Text Multimodal Learning

The recent rapid advancements in language models (LMs) have garnered attention in medical time series-text multimodal learning. However, existing contrastive learning-based and prompt-based LM approaches tend to be biased, often assigning a primary role to time series modality while treating text modality as secondary. We classify these approaches under a temporal-primary paradigm, which may overlook the unique and critical task-relevant information embedded in text modality like clinical reports, thus failing to fully leverage mutual benefits and complementarity of different modalities. To fill this gap, we propose a novel textual-temporal multimodal learning paradigm that enables either modality to serve as the primary while being enhanced by the other, thereby effectively capturing modality-specific information and fostering cross-modal interaction. In specific, we design MedualTime, a language model composed of dual adapters to implement temporal-primary and textual-primary modeling simultaneously. Within each adapter, lightweight adaptation tokens are injected into the top layers of LM to encourage high-level modality fusion. The shared LM pipeline by dual adapters not only achieves adapter alignment but also enables efficient fine-tuning, reducing computational resources. Empirically, MedualTime demonstrates superior performance on medical data, achieving notable improvements of 8% accuracy and 12% F1 in supervised settings. Furthermore, MedualTime's transferability is validated by few-shot label transfer experiments from coarse-grained to fine-grained medical data. https://github.com/start2020/MedualTime

cs.LG↗

arXiv2025

MedSpaformer: a Transferable Transformer with Multi-granularity Token Sparsification for Medical Time Series Classification

Accurate medical time series (MedTS) classification is essential for effective clinical diagnosis, yet remains challenging due to complex multi-channel temporal dependencies, information redundancy, and label scarcity. While transformer-based models have shown promise in time series analysis, most are designed for forecasting tasks and fail to fully exploit the unique characteristics of MedTS. In this paper, we introduce MedSpaformer, a transformer-based framework tailored for MedTS classification. It incorporates a sparse token-based dual-attention mechanism that enables global context modeling and token sparsification, allowing dynamic feature refinement by focusing on informative tokens while reducing redundancy. This mechanism is integrated into a multi-granularity cross-channel encoding scheme to capture intra- and inter-granularity temporal dependencies and inter-channel correlations, enabling progressive refinement of task-relevant patterns in medical signals. The sparsification design allows our model to flexibly accommodate inputs with variable lengths and channel dimensions. We also introduce an adaptive label encoder to extract label semantics and address cross-dataset label space misalignment. Together, these components enhance the model's transferability across heterogeneous medical datasets, which helps alleviate the challenge of label scarcity. Our model outperforms 13 baselines across 7 medical datasets under supervised learning. It also excels in few-shot learning and demonstrates zero-shot capability in both in-domain and cross-domain diagnostics. These results highlight MedSpaformer's robustness and its potential as a unified solution for MedTS classification across diverse settings.

cs.LG↗

arXiv2025

GAP: Gaussianize Any Point Clouds with Text Guidance

3D Gaussian Splatting (3DGS) has demonstrated its advantages in achieving fast and high-quality rendering. As point clouds serve as a widely-used and easily accessible form of 3D representation, bridging the gap between point clouds and Gaussians becomes increasingly important. Recent studies have explored how to convert the colored points into Gaussians, but directly generating Gaussians from colorless 3D point clouds remains an unsolved challenge. In this paper, we propose GAP, a novel approach that gaussianizes raw point clouds into high-fidelity 3D Gaussians with text guidance. Our key idea is to design a multi-view optimization framework that leverages a depth-aware image diffusion model to synthesize consistent appearances across different viewpoints. To ensure geometric accuracy, we introduce a surface-anchoring mechanism that effectively constrains Gaussians to lie on the surfaces of 3D shapes during optimization. Furthermore, GAP incorporates a diffuse-based inpainting strategy that specifically targets at completing hard-to-observe regions. We evaluate GAP on the Point-to-Gaussian generation task across varying complexity levels, from synthetic point clouds to challenging real-world scans, and even large-scale scenes. Project Page: https://weiqi-zhang.github.io/GAP.

cs.CV↗

arXiv2024

MultiPull: Detailing Signed Distance Functions by Pulling Multi-Level Queries at Multi-Step

Reconstructing a continuous surface from a raw 3D point cloud is a challenging task. Recent methods usually train neural networks to overfit on single point clouds to infer signed distance functions (SDFs). However, neural networks tend to smooth local details due to the lack of ground truth signed distances or normals, which limits the performance of overfitting-based methods in reconstruction tasks. To resolve this issue, we propose a novel method, named MultiPull, to learn multi-scale implicit fields from raw point clouds by optimizing accurate SDFs from coarse to fine. We achieve this by mapping 3D query points into a set of frequency features, which makes it possible to leverage multi-level features during optimization. Meanwhile, we introduce optimization constraints from the perspective of spatial distance and normal consistency, which play a key role in point cloud reconstruction based on multi-scale optimization strategies. Our experiments on widely used object and scene benchmarks demonstrate that our method outperforms the state-of-the-art methods in surface reconstruction.

cs.CV↗

arXiv2024

DiffGS: Functional Gaussian Splatting Diffusion

3D Gaussian Splatting (3DGS) has shown convincing performance in rendering speed and fidelity, yet the generation of Gaussian Splatting remains a challenge due to its discreteness and unstructured nature. In this work, we propose DiffGS, a general Gaussian generator based on latent diffusion models. DiffGS is a powerful and efficient 3D generative model which is capable of generating Gaussian primitives at arbitrary numbers for high-fidelity rendering with rasterization. The key insight is to represent Gaussian Splatting in a disentangled manner via three novel functions to model Gaussian probabilities, colors and transforms. Through the novel disentanglement of 3DGS, we represent the discrete and unstructured 3DGS with continuous Gaussian Splatting functions, where we then train a latent diffusion model with the target of generating these Gaussian Splatting functions both unconditionally and conditionally. Meanwhile, we introduce a discretization algorithm to extract Gaussians at arbitrary numbers from the generated functions via octree-guided sampling and optimization. We explore DiffGS for various tasks, including unconditional generation, conditional generation from text, image, and partial 3DGS, as well as Point-to-Gaussian generation. We believe that DiffGS provides a new direction for flexibly modeling and generating Gaussian Splatting.

cs.CV↗

arXiv2024

Data Imputation from the Perspective of Graph Dirichlet Energy

Data imputation is a crucial task due to the widespread occurrence of missing data. Many methods adopt a two-step approach: initially crafting a preliminary imputation (the "draft") and then refining it to produce the final missing data imputation result, commonly referred to as "draft-then-refine". In our study, we examine this prevalent strategy through the lens of graph Dirichlet energy. We observe that a basic "draft" imputation tends to decrease the Dirichlet energy. Therefore, a subsequent "refine" step is necessary to restore the overall energy balance. Existing refinement techniques, such as the Graph Convolutional Network (GCN), often result in further energy reduction. To address this, we introduce a new framework, the Graph Laplacian Pyramid Network (GLPN). GLPN incorporates a U-shaped autoencoder and residual networks to capture both global and local details effectively. Through extensive experiments on multiple real-world datasets, GLPN consistently outperforms state-of-the-art methods across three different missing data mechanisms. The code is available at https://github.com/liguanlue/GLPN.

cs.LG↗

arXiv2024

UDiFF: Generating Conditional Unsigned Distance Fields with Optimal Wavelet Diffusion

Diffusion models have shown remarkable results for image generation, editing and inpainting. Recent works explore diffusion models for 3D shape generation with neural implicit functions, i.e., signed distance function and occupancy function. However, they are limited to shapes with closed surfaces, which prevents them from generating diverse 3D real-world contents containing open surfaces. In this work, we present UDiFF, a 3D diffusion model for unsigned distance fields (UDFs) which is capable to generate textured 3D shapes with open surfaces from text conditions or unconditionally. Our key idea is to generate UDFs in spatial-frequency domain with an optimal wavelet transformation, which produces a compact representation space for UDF generation. Specifically, instead of selecting an appropriate wavelet transformation which requires expensive manual efforts and still leads to large information loss, we propose a data-driven approach to learn the optimal wavelet transformation for UDFs. We evaluate UDiFF to show our advantages by numerical and visual comparisons with the latest methods on widely used benchmarks. Page: https://weiqi-zhang.github.io/UDiFF.

cs.CV↗

arXiv2023

On Solving Unit Commitment with Alternating Current Optimal Power Flow on GPU

We consider the unit commitment (UC) problem that employs the alternating current optimal power flow (ACOPF) constraints, which is formulated as a mixed-integer nonlinear programming problem and thus challenging to solve in practice. We develop a new scalable algorithm based on alternating direction method of multiplier (ADMM), which enables quickly finding a good feasible solution of the UC-ACOPF problem. Our algorithm employs the component-based decomposition that solves a large number of independent small subproblems, each of which represents an individual grid component (e.g., generator, bus, and transmission line), as well as the UC subproblem. We implement the algorithm in Julia, as part of the existing package \texttt{ExaAdmm.jl}, which can efficiently run on GPUs as well as CPUs. The numerical results are reported by using IEEE test instances.

math.OC↗

arXiv2023

A Co-training Approach for Noisy Time Series Learning

In this work, we focus on robust time series representation learning. Our assumption is that real-world time series is noisy and complementary information from different views of the same time series plays an important role while analyzing noisy input. Based on this, we create two views for the input time series through two different encoders. We conduct co-training based contrastive learning iteratively to learn the encoders. Our experiments demonstrate that this co-training approach leads to a significant improvement in performance. Especially, by leveraging the complementary information from different views, our proposed TS-CoT method can mitigate the impact of data noise and corruption. Empirical evaluations on four time series benchmarks in unsupervised and semi-supervised settings reveal that TS-CoT outperforms existing methods. Furthermore, the representations learned by TS-CoT can transfer well to downstream tasks through fine-tuning.

cs.LG↗

arXiv2023

Robust Attributed Graph Alignment via Joint Structure Learning and Optimal Transport

Graph alignment, which aims at identifying corresponding entities across multiple networks, has been widely applied in various domains. As the graphs to be aligned are usually constructed from different sources, the inconsistency issues of structures and features between two graphs are ubiquitous in real-world applications. Most existing methods follow the ``embed-then-cross-compare'' paradigm, which computes node embeddings in each graph and then processes node correspondences based on cross-graph embedding comparison. However, we find these methods are unstable and sub-optimal when structure or feature inconsistency appears. To this end, we propose SLOTAlign, an unsupervised graph alignment framework that jointly performs Structure Learning and Optimal Transport Alignment. We convert graph alignment to an optimal transport problem between two intra-graph matrices without the requirement of cross-graph comparison. We further incorporate multi-view structure learning to enhance graph representation power and reduce the effect of structure and feature inconsistency inherited across graphs. Moreover, an alternating scheme based algorithm has been developed to address the joint optimization problem in SLOTAlign, and the provable convergence result is also established. Finally, we conduct extensive experiments on six unsupervised graph alignment datasets and the DBP15K knowledge graph (KG) alignment benchmark dataset. The proposed SLOTAlign shows superior performance and strongest robustness over seven unsupervised graph alignment methods and five specialized KG alignment methods.

cs.DB↗

arXiv2023

Quantifying Space-Time Load Shifting Flexibility in Electricity Markets

The power grid is undergoing significant restructuring driven by the adoption of wind/solar power and the incorporation of new flexible technologies that can shift load in space and time (e.g., data centers, battery storage, and modular manufacturing). Load shifting is needed to mitigate space-time fluctuations associated with wind/solar power and other disruptions (e.g., extreme weather). The impact of load shifting on electricity markets is typically quantified via sensitivity analysis, which aims to assess impacts in terms of price volatility and total welfare. This sensitivity approach does not explicitly quantify operational flexibility (e.g., range or probability of feasible operation). In this work, we present a computational framework to enable this; specifically, we quantify operational flexibility by assessing how much uncertainty in net loads (which capture uncertain power injections/withdrawals) can be tolerated by the system under varying levels of load shifting capacity. The proposed framework combines optimization formulations that quantify operational flexibility with power grid models that capture load shifting in the form of virtual links (pathways that transfer load across space-time). Our case studies reveal that adding a single virtual link that shifts load in either space or time can lead to dramatic improvements in system-wide flexibility; this is because shifting relieves space-time congestion that results from transmission constraints and generator ramping constraints. Our results provide insights into how the incorporation of flexible technologies can lead to non-intuitive, system-wide gains in flexibility.

eess.SY↗

arXiv2022

Exploring the Impacts of Power Grid Signals on Data Center Operations using a Receding-Horizon Scheduling Model

Data centers (DCs) can help decarbonize the power grid by helping absorb renewable power (e.g., wind and solar) due to their ability to shift power loads across space and time. However, to harness such load-shifting flexibility, it is necessary to understand how grid signals (carbon signals and market price/load allocations) affect DC operations. An obstacle that arises here is the lack of computationally-tractable DC operation models that can capture objectives, constraints, and information flows that arise at the interface of DCs and the power grid. To address this gap, we present a receding-horizon resource management model (a mixed-integer programming model) that captures the resource management layer between the DC scheduler and the grid while accounting for logical constraints, different types of objectives, and forecasts of incoming job profiles and of available computing capacity. We use our model to conduct extensive case studies based on public data from Microsoft Azure and MISO. Our studies show that DCs can provide significant temporal load-shifting flexibility that results in reduced carbon emissions and peak demand charges. Models and case studies are shared as easy-to-use Julia code.

eess.SY↗

arXiv2022

Pricing and Remunerating Electricity Storage Flexibility Using Virtual Links

Ambitious renewable portfolio standards motivate the incorporation of energy storage resources (ESR) as sources of flexibility. While the United States government aims to promote ESR participation in electricity markets, work on market designs for properly remunerating ESRs is still lacking. In this paper, we propose a new energy market clearing framework that incorporates ESR systems. The new market design is computationally attractive in that it avoids mixed-integer formulations and formulations with complementarity constraints. Moreover, compared to previous market designs, our market decomposes the operations of ESRs using the concept of virtual links, which capture the transfer of energy across time. The virtual link representation reveals economic incentives available for ESR operations and sheds light into how electricity markets should remunerate ESRs. We also explore the role of ESR physical parameters on market behavior; we show that, while energy and power capacity defines the amount of flexibility each ESR can provide, storage charge/discharge efficiencies play a fundamental role in ESR remuneration and in mitigating market price volatility. We use our proposed framework to analyze the interplay between ESRs and independent system operators (ISOs) and to provide insights into optimal deployment strategies of ESRs in power grids.

eess.SY↗

arXiv2021

Wavelength Conversion Efficiency Enhancement in Modal Phase Matched $χ^{(2)}$ Nonlinear Waveguides

Modal phase matching (MPM) is a widely used phase matching technique in Al$_x$Ga$_{1-x}$As and other $χ^{(2)}$ nonlinear waveguides for efficient wavelength conversions. The use of a non-fundamental spatial mode compensates the material dispersion but also reduces the spatial overlap of the three interacting waves and therefore limits the conversion efficiency. In this work, we develop a technique to increase the nonlinear overlap by modifying the material nonlinearity, instead of the traditional method of optimizing the modal field profiles. This could eliminate the limiting factor of low spatial overlap inherent to MPM and significantly enhance the conversion efficiency. Among the design examples provided, this technique could increase the conversion efficiency by a factor of up to $\sim$290 in an Al$_x$Ga$_{1-x}$As waveguide. We further show that this technique is applicable to all $χ^{(2)}$ material systems that utilize MPM for wavelength conversion.

physics.optics↗

arXiv2021

On the Tightness of the Lagrangian Dual Bound for Alternating Current Optimal Power Flow

We study tightness properties of a Lagrangian dual (LD) bound for the nonconvex alternating current optimal power flow (ACOPF) problem. We show an LD bound that can be computed in a parallel, decentralized manner. Specifically, the proposed approach partitions the network into a set of subnetworks, dualizes the coupling constraints (giving the LD function), and maximizes the LD function with respect to the dual variables of the coupling constraints (giving the desired LD bound). The dual variables that maximize the LD are obtained by using a proximal bundle method. We show that the bound is less tight than the popular semidefinite programming relaxation but as tight as the second-order cone programming relaxation. We demonstrate our developments using PGLib-OPF test instances.

math.OC↗