Author

Fu Ming

275 papers · 1987–2026

1987peak 24 in 20032026

Frequent co-authors

Papers

showing 1–25 of 275 RSS

PCONV: The Missing but Desirable Sparsity in DNN Weight Pruning for Real-time Execution on Mobile Devices

Xiaolong Ma, Fu-Ming Guo, Wei Niu +5 more

cs.LG · September 2019 · cited 209× · arXiv ↗ · more like this ·

Model compression techniques on Deep Neural Network (DNN) have been widely acknowledged as an effective way to achieve acceleration on a variety of platforms, and DNN weight pruning is a straightforward and effective method. There are currently two mainstreams of pruning methods representing two extremes of pruning regularity: non-structured, fine-grained pruning can achieve high sparsity and accuracy, but is not hardware friendly; structured, coarse-grained pruning exploits hardware-efficient structures in pruning, but suffers from accuracy drop when the pruning rate is high. In this paper, we introduce PCONV, comprising a new sparsity dimension, -- fine-grained pruning patterns inside the coarse-grained structures. PCONV comprises two types of sparsities, Sparse Convolution Patterns (SCP) which is generated from intra-convolution kernel pruning and connectivity sparsity generated from…

Data hiding watermarking for halftone images

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · IEEE Transactions on Image Processing · April 2002 · cited 193× · DOI ↗ · more like this ·

In many printer and publishing applications, it is desirable to embed data in halftone images. We proposed some novel data hiding methods for halftone images. For the situation in which only the halftone image is available, we propose data hiding smart pair toggling (DHSPT) to hide data by forced complementary toggling at pseudo-random locations within a halftone image. The complementary pixels are chosen to minimize the chance of forming visually undesirable clusters. Our experimental results suggest that DHSPT can hide a large amount of hidden data while maintaining good visual quality. For the situation in which the original multitone image is available and the halftoning method is error diffusion, we propose the modified data hiding error diffusion (MDHED) that integrates the data hiding operation into the error diffusion process. In MDHED, the error due to the data hiding is…

Reweighted Proximal Pruning for Large-Scale Language Representation

Fu-Ming Guo, Sijia Liu, Finlay S. Mungall +2 more

cs.LG · September 2019 · cited 74× · arXiv ↗ · more like this ·

Recently, pre-trained language representation flourishes as the mainstay of the natural language understanding community, e.g., BERT. These pre-trained language representations can create state-of-the-art results on a wide range of downstream tasks. Along with continuous significant performance improvement, the size and complexity of these pre-trained neural models continue to increase rapidly. Is it possible to compress these large-scale language representation models? How will the pruned language representation affect the downstream multi-task transfer learning objectives? In this paper, we propose Reweighted Proximal Pruning (RPP), a new pruning method specifically designed for a large-scale language representation model. Through experiments on SQuAD and the GLUE benchmark suite, we show that proximal pruned BERT keeps high accuracy for both the pre-training task and the downstream…

Data hiding in halftone images by stochastic error diffusion

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 13 Nov 2002 · cited 70× · DOI ↗ · more like this ·

We propose a novel method called DHSED (data hiding stochastic error diffusion) to hide binary visual patterns in two error diffused halftone images. While one halftone image is only a regular error diffused image, stochastic error diffusion is applied to the other image to generate special stochastic characteristics with respect to the first image such that the visual pattern would appear when the two halftone images are overlaid Simulation results show that the two halftone images have good visual quality, and the hidden pattern appears with "normal" and "lower-than-normal" intensity when the two halftone images are overlaid.

<title>Data hiding for halftone images</title>

Fu Ming, Oscar C. L. Au

Computer Vision and Pattern Recognition · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 9 May 2000 · cited 56× · DOI ↗ · more like this ·

With the ease of distribution of digital images, there is a growing concern for copyright control and authentication. While there are many existing watermarking and data hiding methods for natural images, almost none can be applied to halftone images. In this paper, we proposed two novel data hiding methods for halftone images. The proposed Data Hiding Pair-Toggling (DHPT) hides data by forced complementary toggling at pseudo-random locations within a halftone image. It is found to be very effective for halftone images with relatively coarse textures. For halftone images with fine textures (such as error diffusion with Steinberg kernel), the proposed Data Hiding Error Diffusion (DHED) gives significantly better visual quality by integrating the data hiding into the error diffusion operation. Both DHPT and DHED are computationally very simple and yet effective in hiding a relatively…

A Quality-of-Content-Based Joint Source and Channel Coding for Human Detections in a Mobile Surveillance Cloud

Xiang Chen, Jenq–Neng Hwang, De Meng +3 more

Computer Vision and Pattern Recognition · IEEE Transactions on Circuits and Systems for Video Technology · 8 Mar 2016 · cited 55× · DOI ↗ · more like this ·

More than 70% of consumer mobile Internet traffic will be mobile video transmissions by 2019. The development of wireless video transmission technologies has been boosted by the rapidly increasing demand of video streaming applications. Although more and more videos are delivered for video analysis (e.g., object detection/tracking and action recognition), most existing wireless video transmission schemes are developed to optimize human perception quality and are suboptimal for video analysis. In mobile surveillance networks, a cloud server collects videos from multiple moving cameras and detects suspicious persons in all camera views. Camera mobility in smartphones or dash cameras implies that video is to be uploaded through bandwidth-limited and error-prone wireless networks, which may cause quality degradation of the decoded videos and jeopardize the performance of video analyses. In…

Research on Job Satisfaction of Elementary and High School Teachers and Strategies to Increase Job Satisfaction

XU Fu-ming, Jiliang Shen

Education · Chinese Education & Society · September 2007 · cited 50× · DOI ↗ · more like this ·

Job satisfaction is an important topic in teacher education research. Exploring the issue of teacher job satisfaction helps us gain a deeper understanding of teachers' mental state, such as their occupational attitudes, zeal for teaching, and work enthusiasm, which affects the quality of teaching and education. From an examination of teachers' job satisfaction studies conducted in China (including Hong Kong) in the past ten years, we identified changes in job satisfaction among China's teachers and proposed suggestions that school and educational administrators may find relevant in their decision making and administration.

Joint visual cryptography and watermarking

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 21 Mar 2005 · cited 42× · DOI ↗ · more like this ·

In this paper, we discuss how to use the watermarking technique for visual cryptography. Both halftone watermarking and visual cryptography involve a hidden secret image. However, their concepts are different. For visual cryptography, a set of shared binary images is used to protect the content of the hidden image. The hidden image can only be revealed when enough shared images are obtained. For watermarking, the hidden image is usually embedded in a single halftone image while preserving the quality of the watermarked halftone image. In this paper, we propose a joint visual-cryptography and watermarking (JVW) algorithm that has the merits of both visual cryptography and watermarking

A new pheromone updating strategy in ant colony optimization

Jun Sun, Shengwu Xiong, Fu-Ming Guo

Artificial Intelligence · 22 Feb 2005 · cited 34× · DOI ↗ · more like this ·

This work presents a new pheromone updating strategy , which is used to optimize ACO (ant colony optimization) in solving the traveling salesman problem. At first, the paper introduces the principle, the characteristics, the construction and the realization method about the ACO. Then, an improved ant colony optimization algorithm using a new pheromone updating strategy is proposed. The pheromone trail of each edge is set with a lower limit at the beginning iterations of the algorithm, and the worst ant judged by its tour length like the best ant used in ACO is allowed to perform global trail updating. At last, we demonstrate the efficiency of the algorithm by means of experimental study.

Hiding data in halftone image using modified data hiding error diffusion

Fu Ming, Oscar C. L. Au

Computer Vision and Pattern Recognition · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 30 May 2000 · cited 29× · DOI ↗ · more like this ·

A Multi Time Scale Wind Power Forecasting Model of a Chaotic Echo State Network Based on a Hybrid Algorithm of Particle Swarm Optimization and Tabu Search

Xiaomin Xu, Dongxiao Niu, Fu Ming +2 more

Artificial Intelligence · Energies · 3 Nov 2015 · cited 22× · DOI ↗ · more like this ·

The uncertainty and regularity of wind power generation are caused by wind resources’ intermittent and randomness. Such volatility brings severe challenges to the wind power grid. The requirements for ultrashort-term and short-term wind power forecasting with high prediction accuracy of the model used, have great significance for reducing the phenomenon of abandoned wind power , optimizing the conventional power generation plan, adjusting the maintenance schedule and developing real-time monitoring systems. Therefore, accurate forecasting of wind power generation is important in electric load forecasting. The echo state network (ESN) is a new recurrent neural network composed of input, hidden layer and output layers. It can approximate well the nonlinear system and achieves great results in nonlinear chaotic time series forecasting. Besides, the ESN is simpler and less computationally…

A robust public watermark for halftone images

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 25 Jun 2003 · cited 20× · DOI ↗ · more like this ·

In many printing and publishing applications, it is desirable to embed data in halftone images for copyright control and authentication purposes. While intentional attacks on printed matter may not be likely, unintentional attacks such as cropping and distortion due to dirt or human writing/marking are likely. In this paper, we proposed a novel halftone image watermarking method called Watermarking Error Diffusion (WED) to embed a robust, invisible, public watermark in the parity domain of halftone images during halftoning while introducing minimal distortion. Since WED uses the local stochastic properties of the halftone image to embed the watermark throughout the whole image, WED is robust to cropping and distortion due to dirt or human writing/marking. Experimental results suggest that WED can give halftone images with good visual quality and the watermark is robust to crop and human…

Design of AXI bus based MPSoC on FPGA

Fu-ming Xiao, Dongsheng Li, Gaoming Du +3 more

Computer Networks and Communications · August 2009 · cited 17× · DOI ↗ · more like this ·

While the computational core is becoming faster and faster, the communication efficiency between the processors has become a bottleneck which limits the performance of multiprocessor system-on-chip (MPSoC). This paper focuses on design and implementation of AXI bus protocol-based MPSoC architecture. Firstly, the RTL models of 4 NIOS II processors using AXI communication architecture are developed. Then the MPSoC was implemented in Altera Stratix II EP2S180 FPGA. Lastly, the performance was evaluated using matrix operation benchmark and compared with previous in-house designed architecture. Experiments showed that the proposed prototype could run at 100 MHz requiring 8963 adaptive look-up table (ALUTs) and the maxim speedup ratio can be up to 3.81, and performs better than the traditional bus (AHB bus) and 2-D mesh NoC architecture.

A multi-stage collaborative filtering approach for mobile recommendation

Li-Hua Li, Fu-Ming Lee, Yu‐Chun Chen +1 more

Information Systems · 15 Feb 2009 · cited 15× · DOI ↗ · more like this ·

Location-based and personalized services are the key factors for promoting user satisfaction. However, most service providers did not consider the needs of mobile user in terms of their location and event-participation. Consequently, the service provider may lose the chance for better service and profit. In this paper, we present a Multi-stage Collaborative Filtering (MSCF) process to provide event recommendation based on mobile user's location. To achieve this purpose, the Collaborative Filtering (CF) technique is employed and the Adaptive Resonance Theory (ART) network is applied to cluster mobile users according to their personal profile. Sequential pattern mining is, then, used to discover the correlations between events for recommendation. The MSCF is designed not only to recommend for the old registered mobile user (ORMU), but also to handle the cold-start problem for new…

Watermarking technique for color halftone images

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 28 Sep 2004 · cited 14× · DOI ↗ · more like this ·

In this paper, we propose a method to hide invisible patterns in color error diffused halftone images. The hidden pattern is embedded in different color components. The hidden patterns would be revealed when the watermarked color halftone images are under Boolean operation or overlaid. Simulation results show that the watermarked color halftone images have good visual quality and the hidden pattern is visible clearly.

A novel method to embed watermark in different halftone images: data hiding by conjugate error diffusion (DHCED)

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 2003 · cited 14× · DOI ↗ · more like this ·

In this paper, we propose a novel way called DHCED to hide invisible patterns in two or more visually different halftone images (e.g. Lena and Harbor) such that the hidden patterns would appear on the halftone images when they are overlaid. Conjugate error diffusion is used to embed the binary visual pattern in the two distinct halftone images. Simulation results show that the two halftone images have good visual quality, and the hidden pattern is visible when the two distinct halftone images are overlaid.

Improved halftone image data hiding with intensity selection

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 13 Nov 2002 · cited 13× · DOI ↗ · more like this ·

In this paper, we propose a novel algorithm called intensity selection (IS) that can be applied to three existing halftone image data hiding algorithms DHST, DHPT and DHSPT to achieve improved visual quality. The proposed IS algorithm generalizes the hidden data representation and select the best location out of a set of candidate locations for the application of DHST/DHPT/DHSPT. It chooses pixel locations that are either very bright or very dark. The IS requires the inverse-halftoned image which implies potentially high computation requirement. Experiments suggest that significant improvement in visual quality can be achieved.

Formality based genetic programming

Pei He, Lishan Kang, Fu Ming

Artificial Intelligence · June 2008 · cited 10× · DOI ↗ · more like this ·

Genetic programming (GP) is an illogical method for automatic programming. It shows creativity in discovering a desired program to solve problem, but in essence bases its searching principle on software testing. This paper is dedicated to establishing a novel GP which combines classical GP and formal approaches like Hoare’s logic, model checking, and automaton, etc. The result indicates these methods can collaborate in the framework pretty well. As has been demonstrated by the experiment, they work in a way that preserves their advantages while each compensates for the deficiencies of the other. So, once an approximate program is obtained, we can say with certainty it is correct with respect to its corresponding pre- and post-conditions.

Review of Recommender Systems and Their Applications

Li-Hua Li, Rong-Wang Hsu, Fu-ming Lee

Information Systems · June 2012 · cited 9× · DOI ↗ · more like this ·

In the past few decades, the concepts of personalization in the forms of recommender system, information filtering, or customization are not only quickly becoming more accepted by the public, but is also drawing considerable attention from enterprises. Therefore, a number of studies based on personalized recommendations have subsequently been produced. Most of these studies apply to E-commerce, website, and information, and some apply to teaching, tourism, and TV programs. Because of the recent rise of Web 3.0 and emphasis on providing more complete personal information and service through an efficient method, the recommender application has gradually developed towards mobile commerce, mobile information, or social network. Many studies have adopted Content- Based (CB), Collaborative Filtering (CF), and the hybrid approach as the main recommender style in the analysis. There are few or…

Correlation-based watermarking for halftone images

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 4 Oct 2004 · cited 9× · DOI ↗ · more like this ·

In many printer and publishing applications, it is desirable to embed data in halftone images for copyright control and authentication purposes. While intentional attacks on printed matters may not be likely, unintentional attacks such as cropping and distortion due to dirt or human writing/marking are likely. In this paper, we propose a novel halftone image watermarking method called Correlation-based Watermarking Error Diffusion (CWED) to embed a full-size, high-resolution and robust binary image in a single halftone image during halftoning while introducing minimal distortion. Since CWED uses the stochastic properties of halftone images to embed the watermark throughout the whole image, CWED is robust to cropping and distortion due to dirt or marking.

The first order load-balanced algorithm with static fixing scheme for centralized WSN system in outdoor environmental monitoring

Yijing Chu, Chu-Ping Tseng, K.-C. Liao +7 more

Computer Networks and Communications · October 2009 · cited 8× · DOI ↗ · more like this ·

The power of nodes in the first layer of a wireless sensor network (WSN) cannot last due to their frequent relay tasks. One possible solution to this disadvantage is to adopt the first order load-balanced algorithm (LBA) to redistribute the loading of the first layer node uniformly. The balanced low-latency convergecast tree (BLLCT) algorithm, a LBA, is generally utilized in route planning in ecological surveillance. Given that the balance between layers does not ensure the balance of the whole WSN, we propose a LBA combining with a static fixing scheme to shift the node groups from the subtree with the largest loading to the one with smaller loading. This procedures repeats until the loading of each subtree is approximately identical. Such a modified LBA successfully overcomes many complicated problems when a WSN is implemented, based on the simulation results using grid and random…

The Blog-article Recommendation System(BARS)

Li-Hua Li, Fu-Ming Lee, Shang-Chi Chan

Information Systems · 2008 · cited 8× · OpenAlex ↗ · more like this ·

blog sites is up rising to 72 millions and the popularity of blogers has drawn many attention. This phenomenon has turned many web users into bloggers. The vast amount of blog information also brings the phenomenon of information overloading which is not handled by the blog function yet. In addition, the personalized recommendation service, which should be provided, is also not incorporated in the blog function now. To better service the bloggers and to overcome the above problems, this research proposes a Blog Article Recommendation System (BARS) which provides personalized article recommendation based on blogger’s interests. This research adapts the ontology technique in BARS to construct a personal preference tree for understanding blogger’s interests. The ART (Adaptive Resonance Theory) network is also utilized to cluster the group with similar interests. In order to find the…

Data hiding in halftone images by conjugate error diffusion

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 31 Oct 2003 · cited 8× · DOI ↗ · more like this ·

In this paper, we propose a method called Data Hiding by Conjugate Error Diffusion (DHCED) to hide invisible patterns in two or more error diffused halftone images such that the hidden patterns would appear on the halftone images when they are overlaid. A binary visual pattern is hidden in two halftone images by applying Conjugate Error Diffusion. Simulation results show that the two halftone images have good visual quality, and the hidden pattern is visible when the two halftone images are overlaid. Compared with the old proposed method, Data Hiding by Stochastic Error Diffusion (DHSED), the contrast between the hidden pattern and the background is greatly improved. The minor trace of edge generated by the phase change of DHSED is also eliminated.

QoS-Based Learning Services Composition for Ubiquitous Learning

Fu-Ming Huang, Ci-Wei Lan, Stephen J.H. Yang

Information Systems · June 2008 · cited 7× · DOI ↗ · more like this ·

With the advance of Internet technologies and handheld computer, lots of software components are built for ubiquitous learning activities. Meanwhile, many studies devote to the promotion of a universal access way and wider reusability of these ubiquitous learning components with the Web services. Nevertheless, the QoS capability will be a decisive characteristic to distinguish services with identical functionalities from each other. In this paper, we propose an efficient service selection scheme to help teacher pick out learning services by considering two different contexts namely single QoS-based service discovery and QoS-based optimization of service composition. Based on different concerns, we present the corresponding solutions to demonstrate the evaluation process of QoS capability for service selection. The experimental results not only showed our solution is more efficient than…

A novel self conjugate halftone image watermarking technique

Fu Ming, Oscar C. Au

Computer Vision and Pattern Recognition · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 21 Nov 2003 · cited 7× · DOI ↗ · more like this ·

In this paper, we propose a novel way called Data Hiding Self Conjugate Error Diffusion to hide invisible patterns in a single error diffused halftone image. The hidden patterns would appear on the halftone image when the halftone image is folded or overlaps with its own rotated version. Simulation results show that the halftone images have good visual quality and the hidden pattern is visible clearly.