Author name cluster

Chuang Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

10 papers

2 author rows

EAAI Journal 2025 Journal Article

Control oriented fast optimisation with SHapley additive explanations assisted two-stage training of input convex neural network

Chuang Wang
Lijun Zhang

Advancement of technology has enabled operation optimisation control of complex systems, which involves solving an optimisation problem. However, there are still challenges to harness the power of such technologies. This is especially true when it is required to solve a complex optimisation problem within prescribed time and resource limits. This paper introduces a novel fast optimisation framework by utilising input convex neural networks to approximate time-consuming optimal control algorithms. A two-stage training structure, consisting of offline data training and online fine-tuning, is put forward to ensure the effectiveness of the trained model. Moreover, an expansion data collection method and a roaming-style data augmentation strategy are proposed to improve the quality of training samples. A SHapley additive explanations assisted data selection algorithm is then developed to identify the most useful samples for training the model. Thus, reducing the training complexity. Further, a predictive safety filter is employed to guarantee that the control inputs produced by the trained model do not violate any constraints. Experiments on an inverted pendulum problem under four types of disturbances demonstrated that the trained input convex neural network controller exhibits superior control performance in comparison to a model predictive controller. Experimental results indicate that, in benchmark inverted pendulum on a cart control tests encompassing eight distinct scenarios, the trained input convex neural network controller achieves similar control performance comparable to that of the model predictive controller. Compared with the time model predictive controller, the trained input convex neural network controller demonstrates superior performance in scenarios with large initial angles and disturbances, though it performs slightly worse under disturbance-free conditions. Furthermore, the trained input convex neural network controller significantly outperforms both the model predictive controller and time model predictive controller in terms of computational efficiency, reducing computation time by 49. 42% and 94. 71%, respectively.

Details DOI

AAAI Conference 2025 Conference Paper

TrackGo: A Flexible and Efficient Method for Controllable Video Generation

Haitao Zhou
Chuang Wang
Rui Nie
Jinlin Liu
Dongdong Yu
Qian Yu
Changhu Wang

Recent years have seen substantial progress in diffusion-based controllable video generation. However, achieving precise control in complex scenarios, including fine-grained object parts, sophisticated motion trajectories, and coherent background movement, remains a challenge. In this paper, we introduce *TrackGo*, a novel approach that leverages free-form masks and arrows for conditional video generation. This method offers users with a flexible and precise mechanism for manipulating video content. We also propose the *TrackAdapter* for control implementation, an efficient and lightweight adapter designed to be seamlessly integrated into the temporal self-attention layers of a pretrained video generation model. This design leverages our observation that the attention map of these layers can accurately activate regions corresponding to motion in videos. Our experimental results demonstrate that our new approach, enhanced by the TrackAdapter, achieves state-of-the-art performance on key metrics such as FVD, FID, and ObjMC scores.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

ViewCraft3D: High-fidelity and View-Consistent 3D Vector Graphics Synthesis

Chuang Wang
Haitao Zhou
Ling Luo
Qian Yu

3D vector graphics play a crucial role in various applications including 3D shape retrieval, conceptual design, and virtual reality interactions due to their ability to capture essential structural information with minimal representation. While recent approaches have shown promise in generating 3D vector graphics, they often suffer from lengthy processing times and struggle to maintain view consistency. To address these limitations, we propose VC3D ( V iew C raft 3D ), an efficient method that leverages 3D priors to generate 3D vector graphics. Specifically, our approach begins with 3D object analysis, employs a geometric extraction algorithm to fit 3D vector graphics to the underlying structure, and applies view-consistent refinement process to enhance visual quality. Our comprehensive experiments demonstrate that VC3D outperforms previous methods in both qualitative and quantitative evaluations, while significantly reducing computational overhead. The resulting 3D sketches maintain view consistency and effectively capture the essential characteristics of the original objects.

PDF Details

NeurIPS Conference 2024 Conference Paper

Latent Neural Operator for Solving Forward and Inverse PDE Problems

Tian Wang
Chuang Wang

Neural operators effectively solve PDE problems from data without knowing the explicit equations, which learn the map from the input sequences of observed samples to the predicted values. Most existing works build the model in the original geometric space, leading to high computational costs when the number of sample points is large. We present the Latent Neural Operator (LNO) solving PDEs in the latent space. In particular, we first propose Physics-Cross-Attention (PhCA) transforming representation from the geometric space to the latent space, then learn the operator in the latent space, and finally recover the real-world geometric space via the inverse PhCA map. Our model retains flexibility that can decode values in any position not limited to locations defined in the training set, and therefore can naturally perform interpolation and extrapolation tasks particularly useful for inverse problems. Moreover, the proposed LNO improves both prediction accuracy and computational efficiency. Experiments show that LNO reduces the GPU memory by 50%, speeds up training 1. 8 times, and reaches state-of-the-art accuracy on four out of six benchmarks for forward problems and a benchmark for inverse problem. Code is available at https: //github. com/L-I-M-I-T/LatentNeuralOperator.

PDF Details DOI

NeurIPS Conference 2023 Conference Paper

DiffSketcher: Text Guided Vector Sketch Synthesis through Latent Diffusion Models

XiMing Xing
Chuang Wang
Haitao Zhou
Jing Zhang
Qian Yu
Dong Xu

Even though trained mainly on images, we discover that pretrained diffusion models show impressive power in guiding sketch synthesis. In this paper, we present DiffSketcher, an innovative algorithm that creates \textit{vectorized} free-hand sketches using natural language input. DiffSketcher is developed based on a pre-trained text-to-image diffusion model. It performs the task by directly optimizing a set of Bézier curves with an extended version of the score distillation sampling (SDS) loss, which allows us to use a raster-level diffusion model as a prior for optimizing a parametric vectorized sketch generator. Furthermore, we explore attention maps embedded in the diffusion model for effective stroke initialization to speed up the generation process. The generated sketches demonstrate multiple levels of abstraction while maintaining recognizability, underlying structure, and essential visual details of the subject drawn. Our experiments show that DiffSketcher achieves greater quality than prior work. The code and demo of DiffSketcher can be found at https: //ximinng. github. io/DiffSketcher-project/.

PDF Details

AAAI Conference 2021 Conference Paper

Proxy Graph Matching with Proximal Matching Networks

Hao-Ru Tan
Chuang Wang
Si-Tong Wu
Tie-Qiang Wang
Xu-Yao Zhang
Cheng-Lin Liu

Estimating feature point correspondence is a common technique in computer vision. A line of recent data-driven approaches utilizing the graph neural networks improved the matching accuracy by a large margin. However, these learning-based methods require a lot of labeled training data, which are expensive to collect. Moreover, we find most methods are sensitive to global transforms, for example, a random rotation. On the contrary, classical geometric approaches are immune to rotational transformation though their performance is generally inferior. To tackle these issues, we propose a new learning-based matching framework, which is designed to be rotationally invariant. The model only takes geometric information as input. It consists of three parts: a graph neural network to generate a high-level local feature, an attention-based module to normalize the rotational transform, and a global feature matching module based on proximal optimization. To justify our approach, we provide a convergence guarantee for the proximal method for graph matching. The overall performance is validated by numerical experiments. In particular, our approach is trained on the synthetic random graphs and then applied to several real-world datasets. The experimental results demonstrate that our method is robust to rotational transform and highlights its strong performance of matching accuracy.

PDF Details

NeurIPS Conference 2019 Conference Paper

A Solvable High-Dimensional Model of GAN

Chuang Wang
Hong Hu
Yue Lu

We present a theoretical analysis of the training process for a single-layer GAN fed by high-dimensional input data. The training dynamics of the proposed model at both microscopic and macroscopic scales can be exactly analyzed in the high-dimensional limit. In particular, we prove that the macroscopic quantities measuring the quality of the training process converge to a deterministic process characterized by an ordinary differential equation (ODE), whereas the microscopic states containing all the detailed weights remain stochastic, whose dynamics can be described by a stochastic differential equation (SDE). This analysis provides a new perspective different from recent analyses in the limit of small learning rate, where the microscopic state is always considered deterministic, and the contribution of noise is ignored. From our analysis, we show that the level of the background noise is essential to the convergence of the training process: setting the noise level too strong leads to failure of feature recovery, whereas setting the noise too weak causes oscillation. Although this work focuses on a simple copy model of GAN, we believe the analysis methods and insights developed here would prove useful in the theoretical understanding of other variants of GANs with more advanced training algorithms.

PDF Details

JBHI Journal 2017 Journal Article

Automatic Choroidal Layer Segmentation Using Markov Random Field and Level Set Method

Chuang Wang
Ya Xing Wang
Yongmin Li

The choroid is an important vascular layer that supplies oxygen and nourishment to the retina. The changes in thickness of the choroid have been hypothesized to relate to a number of retinal diseases in the pathophysiology. In this paper, an automatic method is proposed for segmenting the choroidal layer from macular images by using the level set framework. The three-dimensional nonlinear anisotropic diffusion filter is used to remove all the optical coherence tomography (OCT) imaging artifacts including the speckle noise and to enhance the contrast. The distance regularization and edge constraint terms are embedded into the level set method to avoid the irregular and small regions and keep information about the boundary between the choroid and sclera. Besides, the Markov random field method models the region term into the framework by correlating the single-pixel likelihood function with neighborhood information to compensate for the inhomogeneous texture and avoid the leakage due to the shadows cast by the blood vessels during imaging process. The effectiveness of this method is demonstrated by comparing against other segmentation methods on a dataset with manually labeled ground truth. The results show that our method can successfully and accurately estimate the posterior choroidal boundary.

Details DOI

IROS Conference 2017 Conference Paper

Neural network modeling for steering control of an autonomous vehicle

Gowtham Garimella
Joseph Funke
Chuang Wang
Marin Kobilarov

Model-based control of dynamical systems typically requires accurate domain-specific knowledge and specifications of possibly proprietary system components. In the context of autonomous driving, steering actuator dynamics can be difficult to model due to an integrated proprietary power steering control module. While first-principles models derived from physics laws can often approximate the system behavior, it remains generally difficult to capture non-physically derived behavior based on proprietary software algorithms in the power steering system. To overcome this limitation, this work instead employs a recurring neural network to model the steering dynamics of an autonomous vehicle. The resulting model is then integrated into a Nonlinear Model Predictive Control scheme to generate feedforward steering commands for embedded control. The proposed approach is compared to traditional first-principles steering modeling through on-vehicle experiments and statistical data validation. As a result, it is shown that the neural network model can be automatically generated with less domain-specific knowledge, can predict steering dynamics more accurately, and perform comparably to a high-fidelity first principles model when used for controlling the steering system of a self-driving vehicle.

Details

NeurIPS Conference 2017 Conference Paper

The Scaling Limit of High-Dimensional Online Independent Component Analysis

Chuang Wang
Yue Lu

We analyze the dynamics of an online algorithm for independent component analysis in the high-dimensional scaling limit. As the ambient dimension tends to infinity, and with proper time scaling, we show that the time-varying joint empirical measure of the target feature vector and the estimates provided by the algorithm will converge weakly to a deterministic measured-valued process that can be characterized as the unique solution of a nonlinear PDE. Numerical solutions of this PDE, which involves two spatial variables and one time variable, can be efficiently obtained. These solutions provide detailed information about the performance of the ICA algorithm, as many practical performance metrics are functionals of the joint empirical measures. Numerical simulations show that our asymptotic analysis is accurate even for moderate dimensions. In addition to providing a tool for understanding the performance of the algorithm, our PDE analysis also provides useful insight. In particular, in the high-dimensional limit, the original coupled dynamics associated with the algorithm will be asymptotically “decoupled”, with each coordinate independently solving a 1-D effective minimization problem via stochastic gradient descent. Exploiting this insight to design new algorithms for achieving optimal trade-offs between computational and statistical efficiency may prove an interesting line of future research.

PDF Details