Author

Han Feng

208 papers · 1995–2026

1995peak 17 in 20262026

Frequent co-authors

Papers

showing 1–25 of 208 RSS

Permutation Equivariant Graph Framelets for Heterophilous Graph Learning

Jianfei Li, Ruigang Zheng, Han Feng +2 more

cs.LG · June 2023 · cited 152× · arXiv ↗ · more like this ·

The nature of heterophilous graphs is significantly different from that of homophilous graphs, which causes difficulties in early graph neural network models and suggests aggregations beyond the 1-hop neighborhood. In this paper, we develop a new way to implement multi-scale extraction via constructing Haar-type graph framelets with desired properties of permutation equivariance, efficiency, and sparsity, for deep learning tasks on graphs. We further design a graph framelet neural network model PEGFAN (Permutation Equivariant Graph Framelet Augmented Network) based on our constructed graph framelets. The experiments are conducted on a synthetic dataset and 9 benchmark datasets to compare performance with other state-of-the-art models. The result shows that our model can achieve the best performance on certain datasets of heterophilous graphs (including the majority of heterophilous…

Deeper Insights into Deep Graph Convolutional Networks: Stability and Generalization

Guangrui Yang, Ming Li, Han Feng +1 more

cs.LG · October 2024 · cited 48× · arXiv ↗ · more like this ·

Graph convolutional networks (GCNs) have emerged as powerful models for graph learning tasks, exhibiting promising performance in various domains. While their empirical success is evident, there is a growing need to understand their essential ability from a theoretical perspective. Existing theoretical research has primarily focused on the analysis of single-layer GCNs, while a comprehensive theoretical exploration of the stability and generalization of deep GCNs remains limited. In this paper, we bridge this gap by delving into the stability and generalization properties of deep GCNs, aiming to provide valuable insights by characterizing rigorously the associated upper bounds. Our theoretical results reveal that the stability and generalization of deep GCNs are influenced by certain key factors, such as the maximum absolute eigenvalue of the graph filter operators and the depth of the…

An Analysis of Teacher-Student Interaction Patterns in a Robotics Course for Kindergarten Children: A Pilot Study

Eric Zhi Feng Liu, Chun Hung Lin, Pey Yan Liou +2 more

Computer Science Applications · 2013 · cited 38× · OpenAlex ↗ · more like this ·

Compared with other media, programmable bricks provide children with the opportunity to create their own product and, through this process, to express creative thinking. Studies have found that learning robotics or integrating programming bricks into courses can help to develop students’ problem-solving abilities and enhance their learning performance. This study attempted to develop a one-to-one Topobo robotics course for kindergarten children and to explore teacher-student interaction patterns. This study used a creative thinking spiral as the framework for the Topobo robotics course. The research sample included a five-year-old child and a preschool teacher. Topobo, the programmable bricks, was the main learning tool in this course, and the sequential analysis method was used to identify teacher-student interaction patterns. Based on the frequency of the teacher-student interactions…

MARINE: Multi-relational Network Embeddings with Relational Proximity and Node Attributes

Ming-Han Feng, Chin-Chi Hsu, Cheng–Te Li +2 more

Artificial Intelligence · 13 May 2019 · cited 28× · DOI ↗ · more like this ·

Network embedding aims at learning an effective vector transformation for entities in a network. We observe that there are two diverse branches of network embedding: for homogeneous graphs and for multi-relational graphs. This paper then proposes MARINE, a unified embedding framework for both homogeneous and multi-relational networks to preserve both the proximity and relation information. We also extend the framework to incorporate existing features of nodes in a graph, which can further be exploited for the ensemble of embedding. Our solution possesses complexity linear to the number of edges, which is suitable for large-scale network applications. Experiments conducted on several real-world network datasets, along with applications in link prediction and multi-label classification, exhibit the superiority of our proposed MARINE.

Local industrial policy and productivity: Evidence from China

Jingjing Chen, Han Feng, Hong Zhou

Political Science and International Relations · Contemporary Economic Policy · 6 Oct 2021 · cited 15× · DOI ↗ · more like this ·

Abstract Subnational governments played key roles in China's past rapid growth, and local industrial policy was among their most frequently used tools. Using a plausible comprehensive local industrial policies variable constructed using local Five‐Year Plans and a comprehensive dataset of all medium and large firms in China between 1998 and 2013, this paper attempts to understand whether such interventions are helpful for China's growth. Our results indicate that although local industrial policies attract more low productivity firms, they do achieve their purpose of improving the productivity of the targeted sectors, at firm and city levels, especially in the long run.

Gambling in contests with random initial law

Han Feng, David Hobson

econ.GN · May 2014 · cited 12× · arXiv ↗ · more like this ·

This paper studies a variant of the contest model introduced in Seel and Strack [J. Econom. Theory 148 (2013) 2033-2048]. In the Seel-Strack contest, each agent or contestant privately observes a Brownian motion, absorbed at zero, and chooses when to stop it. The winner of the contest is the agent who stops at the highest value. The model assumes that all the processes start from a common value $x_0>0$ and the symmetric Nash equilibrium is for each agent to utilise a stopping rule which yields a randomised value for the stopped process. In the two-player contest, this randomised value has a uniform distribution on $[0,2x_0]$. In this paper, we consider a variant of the problem whereby the starting values of the Brownian motions are independent, nonnegative random variables that have a common law $μ$. We consider a two-player contest and prove the existence and uniqueness of a symmetric…

Approximation analysis of CNNs from a feature extraction view

Jianfei Li, Han Feng, Ding-Xuan Zhou

cs.LG · October 2022 · cited 10× · arXiv ↗ · more like this ·

Deep learning based on deep neural networks has been very successful in many practical applications, but it lacks enough theoretical understanding due to the network architectures and structures. In this paper we establish some analysis for linear feature extraction by a deep multi-channel convolutional neural networks (CNNs), which demonstrates the power of deep learning over traditional linear transformations, like Fourier, wavelets, redundant dictionary coding methods. Moreover, we give an exact construction presenting how linear features extraction can be conducted efficiently with multi-channel CNNs. It can be applied to lower the essential dimension for approximating a high dimensional function. Rates of function approximation by such deep networks implemented with channels and followed by fully-connected layers are investigated as well. Harmonic analysis for factorizing linear…

Mathematical Modeling of Business Reopening when Facing SARS-CoV-2 Pandemic: Protection, Cost and Risk

Hongyu Miao, Qianmiao Gao, Han Feng +11 more

physics.soc-ph · 12 Jun 2020 · cited 9× · arXiv ↗ · more like this ·

The sudden onset of the coronavirus (SARS-CoV-2) pandemic has resulted in tremendous loss of human life and economy in more than 210 countries and territories around the world. While self-protections such as wearing mask, sheltering in place and quarantine polices and strategies are necessary for containing virus transmission, tens of millions people in the U.S. have lost their jobs due to the shutdown of businesses. Therefore, how to reopen the economy safely while the virus is still circulating in population has become a problem of significant concern and importance to elected leaders and business executives. In this study, mathematical modeling is employed to quantify the profit generation and the infection risk simultaneously from the point of view of a business entity. Specifically, an ordinary differential equation model was developed to characterize disease transmission and…

Reverse Hölder's inequality for spherical harmonics

Feng Dai, Han Feng, Sergey Tikhonov

math.CA · 6 Aug 2014 · cited 9× · arXiv ↗ · more like this ·

This paper determines the sharp asymptotic order of the following reverse Hölder inequality for spherical harmonics $Y_n$ of degree $n$ on the unit sphere $\mathbb{S}^{d-1}$ of $\mathbb{R}^d$ as $n\to \infty$: \[\|Y_n\|_{L^q(\mathbb{S}^{d-1})}\leq C n^{α(p,q)}\|Y_n\|_{L^p(\mathbb{S}^{d-1})},\quad 0<p<q\leq \infty.\] In many cases, these sharp estimates turn out to be significantly better than the corresponding estimates in the Nilkolskii inequality for spherical polynomials. Furthermore, they allow us to improve two recent results on the restriction conjecture and the sharp Pitt inequalities for the Fourier transform on $\mathbb{R}^d$.

A Dynamical System Perspective for Escaping Sharp Local Minima in Equality Constrained Optimization Problems

Han Feng, Haixiang Zhang, Javad Lavaei

Numerical Analysis · 14 Dec 2020 · cited 6× · DOI ↗ · more like this ·

This paper provides a dynamical system perspective on the escape of sharp local minima in constrained optimization problems. The dynamical system view models a perturbed projected first-order optimization algorithm and translates the problem of escaping local minima in constrained optimization problems to that of escaping regions of attraction of the corresponding dynamical system. We develop the notion of biased perturbation and show that it gives a quantitative view of the notion of small regions of attraction that can be escaped. As a counterpart, we explain why the dynamics is stable in a wide region of attraction around a strongly stable equilibrium. Numerical examples are provided to illustrate the usefulness of the developed concepts.

A back-propagation neural network based on a hybrid genetic algorithm and particle swarm optimization for image compression

Han Feng, Man Tang, Jie Qi

Computer Vision and Pattern Recognition · October 2011 · cited 6× · DOI ↗ · more like this ·

In this paper, an improved approach integrating genetic algorithm and adaptive particle swarm optimization with feed forward neural networks for image compression is proposed. The hybrid genetic algorithm with a novel mutation strategy and particle swarm optimization is used to train the neural network to near global optimum weights and thresholds at first. Then the network is trained with gradient descending learning algorithm to obtain the optimal network parameters. Then, the trained network is applied to the image compression. Results show that at the same compression rate the application of optimized neural network in image compression will achieve better image quality compared with the application of traditional neural network.

One kind based on double unidimensional chaos system pictureencryption algorithm

Han Feng-ying, Congxu Zhu

Computer Vision and Pattern Recognition · Computer Engineering and Applications Journal · 2007 · cited 6× · OpenAlex ↗ · more like this ·

Proposes pictureencryption algorithm based on the double unidimensional chaos system.This algorithm uses the double unidimensional chaos system to carry onthe substitution and the replacement transformation to the picture,uses the double chaos system to expand the key space,and increased the key complexity;the substitution and set gathers enhanced theencryption security.the fruit indicates that,thealgorithm calculates speed quick,scrambled text distributedcharacteristic good,the scrambled text dense is sensitive,the bright this algorithm had the high security.

How Are the Seven Major Problems of the Institutional Reform of Local Governments to Be Resolved?

Han Feng

Political Science and International Relations · Chinese Education & Society · July 2002 · cited 5× · DOI ↗ · more like this ·

The institutional reform of local governments has become a hot topic—one of general concern to all quarters in society. Institutional reform is a huge systemic project that involves many different aspects. When formulating and implementing the reform plans, consideration must be taken of many different factors—the cultural environment, social guarantees, the economic system, the political system, and the leadership system. Coordinated action should be taken at all levels while the work is advanced in all directions. Otherwise, if constraints [sic] occur in any one aspect, a growing number of reversals could take place because of hindrances resulting from personnel dispositions, redistribution of interests, and shifts in power, even after the institutions have been reduced in number or merged, and the institutional reform could ultimately fail to achieve its objectives. Hence…

Spherical Image Inpainting with Frame Transformation and Data-driven Prior Deep Networks

Jianfei Li, Chaoyan Huang, Raymond Chan +3 more

eess.IV · 29 Sep 2022 · cited 4× · arXiv ↗ · more like this ·

Spherical image processing has been widely applied in many important fields, such as omnidirectional vision for autonomous cars, global climate modelling, and medical imaging. It is non-trivial to extend an algorithm developed for flat images to the spherical ones. In this work, we focus on the challenging task of spherical image inpainting with deep learning-based regularizer. Instead of a naive application of existing models for planar images, we employ a fast directional spherical Haar framelet transform and develop a novel optimization framework based on a sparsity assumption of the framelet transform. Furthermore, by employing progressive encoder-decoder architecture, a new and better-performed deep CNN denoiser is carefully designed and works as an implicit regularizer. Finally, we use a plug-and-play method to handle the proposed optimization model, which can be implemented…

Analysis of the 3D Application Evaluation System of Landscape Based on Hybrid Cooperation of VR and AI

NIU Lina, Han Feng

Artificial Intelligence · Scientific Programming · 22 Oct 2021 · cited 4× · DOI ↗ · more like this ·

The traditional urban landscape planning and design method cannot integrate the design into the whole planning scene and cannot express the planning effect of multiple schemes and large scenes in an efficient, intuitive, and coordinated way. With the development of computer science and technology, virtual reality technology based on AI is playing an important role in many aspects of urban landscape planning and design. In this paper, on the basis of 3D visualization landscape planning and design, 3D visualization urban landscape modeling is carried out using VR combined with the constructed deep neural network (DNN) to establish a 3D database and virtual scene model of the landscape, and 3D representation of the virtual scene landscape is realized. The designed solution allows the user to grab the object and release it into the workspace. If any placement is needed, the user can pick up…

The development of cracks of compacted expansive soil under dry-wet cycling

LI Han-feng

Artificial Intelligence · Journal of Transport Science and Engineering · 2012 · cited 4× · OpenAlex ↗ · more like this ·

In order to study the development process and evolution of cracks of expansive soil embankment in actual highway engineering,a laboratory simulation tests are designed,the development and the change of cracks of compacted expansive soil with the numbers of dry-wet cycling are obtained,and the crack rates of the samples are described and quantitative analyzed with the graph and image processing of MATLAB.The tests and the analysis results show that the cracks of the expansive soil under the compact condition are fine,and the crack rate is small,the first dry-wet cycling can greatly increase the crack rate of the expansive soil,the width and depth are also increased.However,the effect of the following dry-wet cycles on the crack rates are gradually decreased.This study can provide certain significant guidance for the construction of expansive soil embankment.

An Novel Chaotic Image Encryption Algorithm based on Tangent-Delay Ellipse Reflecting Cavity Map System

Han Feng-ying, Congxu Zhu

Computer Vision and Pattern Recognition · Procedia Engineering · 2011 · cited 4× · DOI ↗ · more like this ·

In order to effectively protect digital image information, to improve the efficiency and security of the encrypted image, many effective and feasible image encryption algorithms have been proposed. The recovery is depended on the recovery value of pixel 8 bits plane; the incorrect from low to high pixel value of plane affect the incorrect of Index of all proportion to the pixel value, resulting in the restoration of image quality. Based on the study of digital image encryption technology, the essay proposes the image encryption algorithm based on TD-RECS chaotic system, and the decryption algorithm against attack on the basis of it.

New image encryption algorithm based on combined high-dimension Chaotic system

Han Feng-ying

Computer Vision and Pattern Recognition · 2009 · cited 4× · OpenAlex ↗ · more like this ·

A new image encryption algorithm based on combined high-dimension chaotic system is presented.Firstly,it uses the generalized cat mapping to diffuse the image,then it uses the Lorenz system to confuse the diffused image.The results demonstrate that the algorithm has good properties of confusion and diffusion.The key space is large enough to resist the brute-force attack.For the encrypted image the distribution of pixel-values has a random-like behavior and the values of adjacent pixels satisfy zero correlation,showing that the proposed scheme is of relatively high security.

Object tracking and classification in aerial videos

Jiangjian Xiao, Hui Cheng, Han Feng +1 more

Computer Vision and Pattern Recognition · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 3 Apr 2008 · cited 4× · DOI ↗ · more like this ·

In the intelligence community, aerial video has become one of the fastest growing data sources and it has been extensively used in intelligence, surveillance, reconnaissance, tactical and security applications. This paper presents a tracking approach to detect moving vehicles and person in such videos taken from aerial platform. In our approach, we combine the layer segmentation approach with background stabilization and post-tracking refinement to reliably detect small moving objects at the relatively low processing speed. For each individual moving object, a corresponding layer is created to maintain an independent appearance and motion model during the tracking process. After the online tracking process, we apply a post-tracking refinement process to link the track fragments into a long consistent track ID to further reduce false alarm and increase detection rate. Furthermore, a…

Recognition of insulator based on yolov5 algorithm

Han Feng, Yong Jiang

Computer Vision and Pattern Recognition · 27 Jul 2021 · cited 3× · DOI ↗ · more like this ·

The pollution flashover accident of insulator strings in substation is the main threat to the safety of power systems, water washing operations on the surface of insulator strings can effectively eliminate this risk, and use intelligent robots to perform water washing tasks instead of manual labor is gradually becoming mainstream. In order to accomplish water washing task, the robot needs to find out the position of the insulator string by using vision sensors. In this paper, we build an insulator strings dataset with 150 images, train and compare the performance differences of four models of Yolov5 for insulator strings detection. Experiments show that when performing the insulator string detection, the difference of detection precision between the four Yolov5 models is small, Yolov5x having 0.02 higher precision than Yolov5s. While the difference in inference speed is large, with…

Intrusion detection analysis by integrating roulette wheel and pseudo-random into back propagation networks

Ruey‐Maw Chen, Chun-Han Feng

Computer Networks and Communications · July 2011 · cited 3× · DOI ↗ · more like this ·

Intrusion detection is a critical component of network security; detection schemes fundamentally use the observed characteristics of network packets as a basis for such determinations. In this study, a cluster center distance method is applied to classify packet type. The cluster center is determined using characteristics of a portion of selected packet data samples prior to detecting. Meanwhile, a well-known back-propagation neural network combined with the roulette wheel selection method and pseudo-random rule are combined with back propagation network (BPN) to determine the intrusion packet type. KDDCUP99 data sets were used as the evaluation packet samples of this experiment. Simulation results demonstrate that roulette wheel selection combined with BPN scheme provides higher detection rate for DoS and R2Lattack packets; BPN with pseudo-random rule can yield higher detection rate…

Efficient methods for the linearization and solution of phase-invariant equations

D. A. Langs, Han Feng-ying

Computational Theory and Mathematics · Acta Crystallographica Section A Foundations of Crystallography · July 1995 · cited 3× · DOI ↗ · more like this ·

This paper describes a linear least-squares procedure, whereby, through quadrupole relationships, the 2 pi integers that linearize sets of unique phase-invariant estimates can be determined. It is subsequently shown that the phase solutions for these linear equations can be obtained, even for basis sets of thousands of phases, without having to either build or invert the full least-squares matrix. The final r.m.s. phase errors achieved by this method can typically be less that 5 or 10 degrees.

In-Context Forcing: Uncovering Context Effects in Autoregressive Video Diffusion

Lingxiao Yang, Liu Liu, Moran Li +4 more

cs.CV · 5 Aug 2026 · cited 2× · arXiv ↗ · more like this ·

Current few-step autoregressive video diffusion models depend on previous fully denoised clean frames as context for all denoising steps of the current frame. However, these clean frames leak excessive local details, which causes the model to take shortcuts, resulting in compromised temporal semantics and dynamics. Inspired by the perspective of diffusion as masking, we explore the impact of noisy contexts on few-step autoregressive generation. Yet, simply applying contexts with the same noise levels provides insufficient guidance, leading to poor temporal consistency. To resolve this dilemma, we introduce In-Context Forcing, a progressive autoregressive paradigm that utilizes contexts with decreasing noise levels. By applying less masking to distant frames and more masking to adjacent ones, this approach provides adaptive guidance, effectively ensuring both robust temporal consistency…

A Hitting Time Analysis for Stochastic Time-Varying Functions with Applications to Adversarial Attacks on Computation of Markov Decision Processes

Ali Yekkehkhany, Han Feng, Donghao Ying +1 more

math.OC · 22 Feb 2023 · cited 2× · arXiv ↗ · more like this ·

Stochastic time-varying optimization is an integral part of learning in which the shape of the function changes over time in a non-deterministic manner. This paper considers multiple models of stochastic time variation and analyzes the corresponding notion of hitting time for each model, i.e., the period after which optimizing the stochastic time-varying function reveals informative statistics on the optimization of the target function. The studied models of time variation are motivated by adversarial attacks on the computation of value iteration in Markov decision processes. In this application, the hitting time quantifies the extent that the computation is robust to adversarial disturbance. We develop upper bounds on the hitting time by analyzing the contraction-expansion transformation appeared in the time-variation models. We prove that the hitting time of the value function in the…

A novel compressed sensing based track before detect algorithm for tracking multiple targets

Jing Liu, Chongzhao Han, Han Feng

Artificial Intelligence · International Conference on Information Fusion · 9 Jul 2013 · cited 2× · OpenAlex ↗ · more like this ·

A novel compressed sensing based track-before-detect (TBD) algorithm is presented in this paper, which is named as CS-TBD algorithm. The proposed CS-TBD algorithm reconstructs the whole radar scenario (direction of arrival (DOA)-Doppler plane) for each range gate at consecutive scans using an improved StOMP algorithm, resulting in a three-dimensional range-DOA-Doppler space. The proposed algorithm then performs temporal tracking in the newly built three-dimensional range-DOA-Doppler space, based on the information from multiple illuminations during each scan, as well as among consecutive scans. In the proposed CS-TBD algorithm, the improved StOMP algorithm together with the temporal tracking, can effectively distinguish true targets from false targets and clutter based on information from multiple illuminations.