Author name cluster

Wenbo Zhao

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.

6 papers

1 author row

EAAI Journal 2025 Journal Article

Neural network adaptive force control for pneumatic polishing end-actuator with external disturbances and full-state constrains

Zhiguo Yang
Jiange Kou
Zhanxin Li
Wenbo Zhao
Yushan Ma
Yixuan Wang
Yan Shi

In pneumatic polishing, the nonlinear, time-varying, and uncertain contact characteristics introduce significant modeling inaccuracies, posing substantial challenges to the realization of precise and robust force control. This paper proposes a neural-network adaptive force control strategy for a pneumatic polishing end-actuator under external disturbances and full-state constraints. To estimate the unmeasurable states and enhance the ant disturbance capability, a composite observer is developed to estimate the internal states and external disturbances in real time. Under the adaptive backstepping design framework, a radial-basis-function–neural-network–based adaptive learning mechanism is employed to approximate the nonlinear uncertainties, and a dynamic surface-control structure is introduced to avoid the complexity explosion in conventional recursive designs. Furthermore, a barrier Lyapunov function is integrated to ensure compliance with the full-state constraints throughout the control process. The convergence of the controller is verified through stability analyses, and the effectiveness and superiority of the control scheme is verified via experiments in four different polishing scenarios. The results show that the proposed control method achieves an average force tracking error less than 0. 07 N and convergence time less than 2. 35 s, showing higher control accuracy, faster transient response, and stronger robustness, than similar control algorithms.

Details DOI

AAAI Conference 2025 Conference Paper

Spatial Annealing for Efficient Few-shot Neural Rendering

Yuru Xiao
Deming Zhai
Wenbo Zhao
Kui Jiang
Junjun Jiang
Xianming Liu

Neural Radiance Fields (NeRF) with hybrid representations have shown impressive capabilities for novel view synthesis, delivering high efficiency. Nonetheless, their performance significantly drops with sparse input views. Various regularization strategies have been devised to address these challenges. However, these strategies either require additional rendering costs or involve complex pipeline designs, leading to a loss of training efficiency. Although FreeNeRF has introduced an efficient frequency annealing strategy, its operation on frequency positional encoding is incompatible with the efficient hybrid representations. In this paper, we introduce an accurate and efficient few-shot neural rendering method named Spatial Annealing regularized NeRF (SANeRF), which adopts the pre-filtering design of a hybrid representation. We initially establish the analytical formulation of the frequency band limit for a hybrid architecture by deducing its filtering process. Based on this analysis, we propose a universal form of frequency annealing in the spatial domain, which can be implemented by modulating the sampling kernel to exponentially shrink from an initial one with a narrow grid tangent kernel spectrum. This methodology is crucial for stabilizing the early stages of the training phase and significantly contributes to enhancing the subsequent process of detail refinement. Our extensive experiments reveal that, by adding merely one line of code, SANeRF delivers superior rendering quality and much faster reconstruction speed compared to current few-shot neural rendering methods. Notably, SANeRF outperforms FreeNeRF on the Blender dataset, achieving 700X faster reconstruction speed.

PDF Details DOI

AAAI Conference 2022 Conference Paper

Local Surface Descriptor for Geometry and Feature Preserved Mesh Denoising

Wenbo Zhao
Xianming Liu
Junjun Jiang
Debin Zhao
Ge Li
Xiangyang Ji

3D meshes are widely employed to represent geometry structure of 3D shapes. Due to limitation of scanning sensor precision and other issues, meshes are inevitably affected by noise, which hampers the subsequent applications. Convolultional neural networks (CNNs) achieve great success in image processing tasks, including 2D image denoising, and have been proven to own the capacity of modeling complex features at different scales, which is also particularly useful for mesh denoising. However, due to the nature of irregular structure, CNNs-based denosing strategies cannot be trivially applied for meshes. To circumvent this limitation, in the paper, we propose the local surface descriptor (LSD), which is able to transform the local deformable surface around a face into 2D grid representation and thus facilitates the deployment of CNNs to generate denoised face normals. To verify the superiority of LSD, we directly feed LSD into the classical Resnet without any complicated network design. The extensive experimental results show that, compared to the state-ofthe-arts, our method achieves encouraging performance with respect to both objective and subjective evaluations.

PDF Details

AAAI Conference 2022 Conference Paper

SpikeConverter: An Efficient Conversion Framework Zipping the Gap between Artificial Neural Networks and Spiking Neural Networks

Fangxin Liu
Wenbo Zhao
Yongbiao Chen
Zongwu Wang
Li Jiang

Spiking Neural Networks (SNNs) have recently attracted enormous research interest since their event-driven and braininspired structure enables low-power computation. In image recognition tasks, the best results achieved by SNN so far utilize ANN-SNN conversion methods that replace activation functions in artificial neural networks (ANNs) with integrate-and-fire neurons. Compared to source ANNs, converted SNNs usually suffer from accuracy loss and require a considerable number of time steps to achieve competitive accuracy. We find that the performance degradation of converted SNN stems from the fact that the information capacity of spike trains in transferred networks is smaller than that of activation values in source ANN, resulting in less information being passed during SNN inference. To better correlate ANN and SNN for better performance, we propose a conversion framework to mitigate the gap between the activation value of source ANN and the generated spike train of target SNN. The conversion framework originates from exploring an identical relation in the conversion and exploits temporal separation scheme and novel neuron model for the relation to hold. We demonstrate almost lossless ANN-SNN conversion using SpikeConverter for a wide variety of networks on challenging datasets including CIFAR-10, CIFAR-100, and ImageNet. Our results also show that SpikeConverter achieves the abovementioned accuracy across different network architectures and datasets using 32X - 512X fewer inference time-steps than state-of-the-art ANN- SNN conversion methods.

PDF Details

TCS Journal 2007 Journal Article

Computational complexity of counting problems on 3-regular planar graphs

Mingji Xia
Peng Zhang
Wenbo Zhao

A variety of counting problems on 3-regular planar graphs are considered in this paper. We give a sufficient condition which guarantees that the coefficients of a homogeneous polynomial can be uniquely determined by its values on a recurrence sequence. This result enables us to use the polynomial interpolation technique in high dimension to prove the #P-completeness of problems on graphs with special requirements. Using this method, we show that #3-Regular Bipartite Planar Vertex Covers is #P-complete. Furthermore, we use Valiant’s Holant Theorem to construct a holographic reduction from it to #2, 3-Regular Bipartite Planar Matchings, establishing the #P-completeness of the latter. Finally, we completely classify the problems #Planar Read-twice 3SAT with different ternary symmetric relations according to their computational complexity, by giving several more applications of holographic reduction in proving the #P-completeness of the corresponding counting problems.

Details DOI

TCS Journal 2006 Journal Article

A network flow approach to the Minimum Common Integer Partition Problem

Wenbo Zhao
Peng Zhang
Tao Jiang

In the k-Minimum Common Integer Partition Problem, abbreviated as k-MCIP, we are given k multisets X 1, …, X k of positive integers, the goal is to find an integer multiset T of the minimum size such that for every i, we can partition each of the integers in X i so that the disjoint (multiset) union of their partitions equals T. This problem has applications in computational molecular biology, in particular, ortholog assignment and DNA hybridization fingerprint assembly. The problem is known to be NP-hard for any k ⩾ 2. In this article, we improve the approximation ratio for k-MCIP by viewing this problem as a flow decomposition problem in some flow network. We show an efficient 0. 5625 k -approximation algorithm, improving upon the previously best known 0. 6139 k -approximation algorithm for this problem.

Details DOI