Resource Type

Journal Article 21

Year

2023 2

2022 2

2021 2

2020 2

2019 3

2018 1

2017 1

2016 1

2013 2

2012 1

2007 2

2004 1

2002 1

open ︾

Keywords

Counter 2

Crowd counting 2

Knowledge distillation 2

ANN 1

Bayesian network 1

Binary matrix multiplication 1

Binary neural networks 1

Body mass index 1

Braid entropy 1

Bus-off hitting time 1

CBC - MAC mode 1

CCM mode 1

CTR mode 1

Chinese cohort 1

Complexity detection 1

Controller area network (CAN) 1

Crowd behavior 1

Crowd scene analysis 1

Density estimation 1

open ︾

Search scope:

排序： Display mode:

Forget less, count better: a domain-incremental self-distillation learning benchmark for lifelong crowd counting Research Article

Jiaqi GAO, Jingqi LI, Hongming SHAN, Yanyun QU, James Z. WANG, Fei-Yue WANG, Junping ZHANG

Frontiers of Information Technology & Electronic Engineering 2023, Volume 24, Issue 2,   Pages 187-202 doi: 10.1631/FITEE.2200380

Abstract: has important applications in public safety and pandemic control. A robust and practical system has to be capable of continuously learning with the newly incoming domain data in real-world scenarios instead of fitting one domain only. Off-the-shelf methods have some drawbacks when handling multiple domains: (1) the models will achieve limited performance (even drop dramatically) among old domains after training images from new domains due to the discrepancies in intrinsic data distributions from various domains, which is called catastrophic forgetting; (2) the well-trained model in a specific domain achieves imperfect performance among other unseen domains because of domain shift; (3) it leads to linearly increasing storage overhead, either mixing all the data for training or simply training dozens of separate models for different domains when new ones are available. To overcome these issues, we investigate a new task in incremental domain training setting called lifelong . Its goal is to alleviate catastrophic forgetting and improve the generalization ability using a single model updated by the incremental domains. Specifically, we propose a self-distillation learning framework as a benchmark (forget less, count better, or FLCB) for lifelong , which helps the model leverage previous meaningful knowledge in a sustainable manner for better to mitigate the forgetting when new data arrive. A new quantitative metric, normalized Backward Transfer (nBwT), is developed to evaluate the forgetting degree of the model in the process. Extensive experimental results demonstrate the superiority of our proposed benchmark in achieving a low catastrophic forgetting degree and strong generalization ability.

Keywords： Crowd counting     Knowledge distillation     Lifelong learning    

HTML PDF Collect

A novel convolutional neural network method for crowd counting Research Articles

Jie-hao Huang, Xiao-guang Di, Jun-de Wu, Ai-yue Chen,18s004055@hit.edu.cn,dixiaoguang@hit.edu.cn

Frontiers of Information Technology & Electronic Engineering 2020, Volume 21, Issue 8,   Pages 1119-1266 doi: 10.1631/FITEE.1900282

Abstract: Crowd , in general, is a challenging task due to the large variation of head sizes in the crowds. Existing methods always use a multi-column convolutional neural network (MCNN) to adapt to this variation, which results in an average effect in areas with different densities and brings a lot of noise to the density map. To address this problem, we propose a new method called the segmentation-aware prior network (SAPNet), which generates a high-quality density map without noise based on a coarse head-segmentation map. SAPNet is composed of two networks, i.e., a foreground-segmentation convolutional neural network (FS-CNN) as the front end and a crowd-regression convolutional neural network (CR-CNN) as the back end. With only the single dot annotation, we generate the ground truth of segmentation masks in heads. Then, based on the ground truth, FS-CNN outputs a coarse head-segmentation map, which helps eliminate the noise in regions without people in the density map. By inputting the head-segmentation map generated by the front end, CR-CNN performs accurate estimation and generates a high-quality density map. We demonstrate SAPNet on four datasets (i.e., ShanghaiTech, UCF-CC-50, WorldExpo’10, and UCSD), and show the state-of-the-art performances on ShanghaiTech part and UCF-CC-50 datasets.

Keywords： Crowd counting     Density estimation     Segmentation prior map     Uniform function    

HTML PDF Collect

Aggregated context network for crowd counting

Si-yue Yu, Jian Pu,51174500148@stu.ecnu.edu.cn,jianpu@fudan.edu.cn

Frontiers of Information Technology & Electronic Engineering 2020, Volume 21, Issue 11,   Pages 1535-1670 doi: 10.1631/FITEE.1900481

Abstract: has been applied to a variety of applications such as video surveillance, traffic monitoring, assembly control, and other public safety applications. Context information, such as perspective distortion and background interference, is a crucial factor in achieving high performance for . While traditional methods focus merely on solving one specific factor, we aggregate sufficient context information into the network to tackle these problems simultaneously in this study. We build a fully convolutional network with two tasks, i.e., main density map estimation and auxiliary . The main task is to extract the multi-scale and spatial context information to learn the density map. The auxiliary task gives a comprehensive view of the background and foreground information, and the extracted information is finally incorporated into the main task by late fusion. We demonstrate that our network has better accuracy of estimation and higher robustness on three challenging datasets compared with state-of-the-art methods.

Keywords： 人群计数；卷积神经网络；密度估计；语义分割；多任务学习    

HTML PDF Collect

Crowd modeling based on purposiveness and a destination-driven analysis method Research Articles

Ning Ding, Weimin Qi, Huihuan Qian,hhqian@cuhk.edu.cn

Frontiers of Information Technology & Electronic Engineering 2021, Volume 22, Issue 10,   Pages 1351-1369 doi: 10.1631/FITEE.2000312

Abstract: This study focuses on the multiphase flow properties of crowd motions. Stability is a crucial forewarning factor for the crowd. To evaluate the behaviors of newly arriving pedestrians and the stability of a crowd, a novel motion structure analysis model is established based on purposiveness, and is used to describe the continuity of pedestrians’ pursuing their own goals. We represent the crowd with self-driven particles using a destination-driven analysis method. These self-driven particles are trackable feature points detected from human bodies. Then we use trajectories to calculate these self-driven particles’ purposiveness and select trajectories with high purposiveness to estimate the common destinations and the inherent structure of the crowd. Finally, we use these common destinations and the crowd structure to evaluate the behavior of newly arriving pedestrians and . Our studies show that the purposiveness parameter is a suitable descriptor for middle-density human crowds, and that the proposed destination-driven analysis method is capable of representing complex crowd motion behaviors. Experiments using synthetic and real data and videos of both human and animal crowds have been conducted to validate the proposed method.

Keywords： 人群建模；智能视频监控；人群稳定性    

HTML PDF Collect

Research on Social Risk of the Massing Crowd in Public Venues

Li Jianfeng,Liu Mao,Sui Xiaolin

Strategic Study of CAE 2007, Volume 9, Issue 6,   Pages 88-93

Abstract:

As the development of cities becomes more quick,  the accidents happened in public venues resulted form massing crowd become more and more.  This paper uses social risk to quantify the risk. On one hand,  in terms of former research outcome,  the paper deems that the occurrences of the accidents in public venues,  in a time span, can be described as Poisson distribution, and then, the quantitative model of accident occurrence probabilities can be reasonal out. On the other hand, through the sum of the occurrences of accidents of different severity level, which is to embody the situational probability of accidents of different severity level with its frequencies,  the outcome of accidents mortalities will be obtained,  which is used to figure out the probabilities of the accidents of different casualty numbers. In the end,  the F - N curve will be achieved. To use the F - N curve,  it is able to analyse the social risk of crowd massing venues.  Taking some statistical accidents as references,  it describes how to use the model.  The result proves the model is reasonable and accurate to a certain extent.

Keywords： crowd massing risk     social risk     F-N curve     quantitative risk analysis    

HTML PDF Collect

Low-power, high-speed, and area-efficient sequential circuits by quantum-dot cellular automata: T-latch and counter study Research Article

Mohammad GHOLAMI, Zaman AMIRZADEH

Frontiers of Information Technology & Electronic Engineering 2023, Volume 24, Issue 3,   Pages 457-469 doi: 10.1631/FITEE.2200361

Abstract:

cellular automata (QCA)%29&ck%5B%5D=abstract&ck%5B%5D=keyword'> is a new nanotechnology for the implementation of nano-sized digital circuits. This nanotechnology is remarkable in terms of speed, area, and power consumption compared to complementary metal-oxide-semiconductor (CMOS) technology and can significantly improve the design of various logic circuits. We propose a new method for implementing a in QCA technology in this paper. The proposed method uses the intrinsic features of QCA in timing and clock phases, and therefore, the proposed cell structure is less occupied and less power-consuming than existing implementation methods. In the proposed , compared to previous best designs, reductions of 6.45% in area occupation and 44.49% in power consumption were achieved. In addition, for the first time, a reset-based and a with set and reset capabilities are designed. Using the proposed , a new 3-bit is developed which reduces 2.14% cell numbers compared to the best of previous designs. Moreover, based on the 3-bit , a 4-bit is designed, which reduces 0.51% cell numbers and 4.16% cross-section area compared to previous designs. In addition, two selective s are introduced to count from 0 to 5 and from 2 to 5. Simulations were performed using and tools in coherence vector engine mode. The proposed circuits are compared with related designs in terms of delay, cell numbers, area, and leakage power.

Keywords： Quantum-dot cellular automata (QCA)     Quantum-dot     T-latch     T-flip-flop     Counter     Selective counter     QCADesigner     QCAPro    

HTML PDF Collect

Controller area network node reliability assessment based on observable node information Article

Lei-ming ZHANG, Long-hao TANG, Yong LEI

Frontiers of Information Technology & Electronic Engineering 2017, Volume 18, Issue 5,   Pages 615-626 doi: 10.1631/FITEE.1601029

Abstract: Controller area network (CAN) based fieldbus technologies have been widely used in networked manufacturing systems. As the information channel of the system, the reliability of the network is crucial to the system throughput, product quality, and work crew safety. However, due to the inaccessibility of the nodes’ internal states, direct assessment of the reliability of CAN nodes using the nodes’ internal error counters is infeasible. In this paper, a novel CAN node reliability assessment method, which uses node’s time to bus-off as the reliability measure, is proposed. The method estimates the transmit error counter (TEC) of any node in the network based on the network error log and the information provided by the observable nodes whose error counters are accessible. First, a node TEC estimation model is established based on segmented Markov chains. It considers the sparseness of the distribution of the CAN network errors. Second, by learning the differences between the model estimates and the actual values from the observable node, a Bayesian network is developed for the estimation updating mechanism of the observable nodes. Then, this estimation updating mechanism is transferred to general CAN nodes with no TEC value accessibility to update the TEC estimation. Finally, a node reliability assessment method is developed to predict the time to reach bus-off state of the nodes. Case studies are carried out to demonstrate the effectiveness of the proposed methodology. Experimental results show that the estimates using the proposed model agree well with actual observations.

Keywords： Controller area network (CAN)     Transmit error counter (TEC)     TEC value estimation     Bayesian network     Bus-off hitting time    

HTML PDF Collect

An efficient counter-based Wallace-tree multiplier with a hybrid full adder core for image blending Research Articles

Ayoub SADEGHI, Nabiollah SHIRI, Mahmood RAFIEE, Mahsa TAHGHIGH

Frontiers of Information Technology & Electronic Engineering 2022, Volume 23, Issue 6,   Pages 950-965 doi: 10.1631/FITEE.2100432

Abstract:

We present a new -based Wallace-tree (CBW) 8×8 . The ‍’‍s s are implemented with a new hybrid (FA) cell, which is based on the (TG) technique. The proposed FA, TG-based AND gate, and hybrid half adder (HA) generate :3 (4≤≤7) digital s with the ability to save at least 50% area occupation. Simulations by 90 nm technology prove the superiority of the proposed FA and digital s under different conditions over the state-of-the-art designs. By using the proposed cells, the CBW exhibits high driving capability, low power consumption, and high speed. The CBW has a 0.0147 mm die area in a pad. The post-layout extraction proves the accuracy of experimental implementation. An mechanism is proposed, in which a direct interface between MATLAB and HSPICE is used to evaluate the presented CBW in image processing applications. The peak signal-to-noise ratio (PSNR) and structural similarity index metric (SSIM) are calculated as image quality parameters, and the results confirm that the presented CBW can be used as an alternative to designs in the literature.

Keywords： Full adder     Transmission gate     Counter     Multiplier     Three-dimensional layout     Image blending    

HTML PDF Collect

Current Status of Blood Lipid Level and Prevention Strategy of Lipid Disorder in Chinese Population

Chen Haozhu,Jin Xuejuan

Strategic Study of CAE 2002, Volume 4, Issue 11,   Pages 1-6

Abstract:

To assess current status of serum lipid levels in healthy inhabitants of China and their secular trend in the past 5 decades. Methods： The authors reviewed the contemporary literature (studies published between 1959 and 2002; bibliographies of reviews and original articles). Articles were included if they were following a standardized protocol for serum collection and lipid examination. Thirty-five articles in different target population from Beijin, Shanghai, Guangzhou, Tianjin and Kunming met inclusion criteria. This paper summarized important results of these studies and also the author´s own study performed in Shanghai on lipid epidemiology. Results： (1)The levels of blood total cholesterol (TC), low density lipoprotein cholesterol (LDL-C) and triglyceride (TG) increased along with the age. The levels of TC, TG and LDL-C among women before fifty were similar to that of among men of the same age, but higher in women than men over fifty. The TC, TG and LDL-C levels were higher in urban area than in rural area. (2) The levels of high density lipoprotein cholesterol (HDL-C) kept stability with aging, and they were lower in men than in women. (3)Secular trend analyses showed that the level of TC, TG and LDL-C increased steadily and significantly in all population, which reached the highest level at the end of 1990s, while HDL-C level decreased. (4) The prevalence of hypercholesterolemia (TC＞ 200 - 220 mg/dL) differed greatly among different population, from 17.8% to 54.9% in men, and from 14.9% to 53.4% in women, respectively. (5) Possible determinants for these changes were alternations in the composition of the diet in Chinese population along with the rise of living standards of the people. Daily intake of dietary fat, animal protein and cholesterol increased, while the proportion of carbohydrate decreased. (6) Coronary heart disease has been become the most common heart disease in China correlating with these changes. Prevention strategies： Primary prevention including： (1) resume healthy diet and have reasonable composition of the diet; (2) have lifestyle modifications and reduce other coronary heart disease risk; (3) recommendation for increasing physical activity. Secondary prevention including： If the patient did not achieve recommended target lipid levels after 3 months of lifestyle modifications (dietary therapy and physical activity), he or she should receive pharmacotherapy for dislipidemia.

Keywords： blood lipids     epidemiology     prevention    

HTML PDF Collect

Identification of important factors influencing nonlinear counting systems Research Article

Xinmin ZHANG, Jingbo WANG, Chihang WEI, Zhihuan SONG,xinminzhang@zju.edu.cn,wangjingbobo@zju.edu.cn,chhwei@zju.edu.cn,songzhihuan@zju.edu.cn

Frontiers of Information Technology & Electronic Engineering 2022, Volume 23, Issue 1,   Pages 123-133 doi: 10.1631/FITEE.2000324

Abstract: Identifying factors that exert more influence on system output from data is one of the most challenging tasks in science and engineering. In this work, a of the (SA-GGPR) model is proposed to identify of the . In SA-GGPR, the GGPR model with Poisson likelihood is adopted to describe the . The GGPR model with Poisson likelihood inherits the merits of nonparametric kernel learning and Poisson distribution, and can handle complex s. Nevertheless, understanding the relationships between model inputs and output in the GGPR model with Poisson likelihood is not readily accessible due to its nonparametric and kernel structure. SA-GGPR addresses this issue by providing a quantitative assessment of how different inputs affect the system output. The application results on a simulated and a real have demonstrated that the proposed SA-GGPR method outperforms several state-of-the-art methods in identification accuracy.

Keywords： Important factors     Nonlinear counting system     Generalized Gaussian process regression     Sensitivity analysis     Steel casting-rolling process    

HTML PDF Collect

Detecting interaction/complexitywithin crowd movements using braid entropy Research Papers

Murat AKPULAT, Murat EKİNCİ

Frontiers of Information Technology & Electronic Engineering 2019, Volume 20, Issue 6,   Pages 849-861 doi: 10.1631/FITEE.1800313

Abstract:

The segmentation of moving and non-moving regions in an image within the field of crowd analysis is a crucial process in terms of understanding crowd behavior. In many studies, similar movements were segmented according to the location, adjacency to each other, direction, and average speed. However, these segments may not in turn indicate the same types of behavior in each region. The purpose of this study is to better understand crowd behavior by locally measuring the degree of interaction/complexity within the segment. For this purpose, the flow of motion in the image is primarily represented as a series of trajectories. The image is divided into hexagonal cells and the finite time braid entropy (FTBE) values are calculated according to the different projection angles of each cell. These values depend on the complexity of the spiral structure that the trajectories generated throughout the movement and show the degree of interaction among pedestrians. In this study, behaviors of different complexities determined in segments are pictured as similar movements on the whole. This study has been tested on 49 different video sequences from the UCF and CUHK databases.

Keywords： Crowd behavior     Motion segmentation     Motion entropy     Crowd scene analysis     Complexity detection     Braid entropy    

HTML PDF Collect

MCDB Monte Carlo dosimetry code system and its applications

Deng Li,Li Gang,Chen Chaobin,Ye Tao

Strategic Study of CAE 2012, Volume 14, Issue 8,   Pages 72-77

Abstract:

MCDB is developed for boron neutron capture therapy (BNCT). This system consists of a medical pre-processor, a dose computation and a post-processor. MCDB automatically produces the input file from CT and MRI image data. In Monte Carlo dose calculation, several accelerated measures, such as the fast track technique, mesh tally matrix and material matrix, are developed. In this paper, we proposed a real model simulated by MCNP and MCDB, respectively. The almost same results as MCNP are achieved. MCDB is faster in computational speed than MCNP.

Keywords： MCDB     fast tracking technique     mesh tally matrix     material matrix    

HTML PDF Collect

Processing and analysis of data from microwave humidity sounder onboard FY-3A satellite

He Jieying,Zhang Shengwei

Strategic Study of CAE 2013, Volume 15, Issue 10,   Pages 47-53

Abstract:

Microwave humidity sounder (MWHS) is one of payloads on the Fengyun-3A (FY-3A) satellite. This paper introduces its structure, operation status and data receiving and processing. The paper constructs an inversion model using artificial neural network (ANN) algorithm, and makes comparison with advanced microwave sounding unit advanced microwave sounding unit-B(AMSU-B). The results demonstrate that the model can be operated successfully. Using the simulated brightness temperatures from MWHS from July to December in 2008 in Beijing, the paper derives water vapor density profiles and gives analysis of root mean square. Meanwhile, the paper focuses on brightness temperature values of different scanning lines when the typhoon comes. The paper demonstrates that FY-3A satellite MWHS can retrieve the water vapor density profiles, cloud liquid water and other related information. Also, in the process of monitoring the tropical typhoon and cyclone and judging the trend of them, FY-3A satellite MWHS also plays a very important role.

Keywords： MWHS     FY-3A     ANN     water vapor density    

HTML PDF Collect

Binary neural networks for speech recognition Regular Papers

Yan-min QIAN, Xu XIANG

Frontiers of Information Technology & Electronic Engineering 2019, Volume 20, Issue 5,   Pages 701-715 doi: 10.1631/FITEE.1800469

Abstract:

Recently, deep neural networks (DNNs) significantly outperform Gaussian mixture models in acoustic modeling for speech recognition. However, the substantial increase in computational load during the inference stage makes deep models difficult to directly deploy on low-power embedded devices. To alleviate this issue, structure sparseness and low precision fixed-point quantization have been applied widely. In this work, binary neural networks for speech recognition are developed to reduce the computational cost during the inference stage. A fast implementation of binary matrix multiplication is introduced. On modern central processing unit (CPU) and graphics processing unit (GPU) architectures, a 5–7 times speedup compared with full precision floatingpoint matrix multiplication can be achieved in real applications. Several kinds of binary neural networks and related model optimization algorithms are developed for large vocabulary continuous speech recognition acoustic modeling. In addition, to improve the accuracy of binary models, knowledge distillation from the normal full precision floating-point model to the compressed binary model is explored. Experiments on the standard Switchboard speech recognition task show that the proposed binary neural networks can deliver 3–4 times speedup over the normal full precision deep models. With the knowledge distillation from the normal floating-point models, the binary DNNs or binary convolutional neural networks (CNNs) can restrict the word error rate (WER) degradation to within 15.0%, compared to the normal full precision floating-point DNNs or CNNs, respectively. Particularly for the binary CNN with binarization only on the convolutional layers, the WER degradation is very small and is almost negligible with the proposed approach.

Keywords： Speech recognition     Binary neural networks     Binary matrix multiplication     Knowledge distillation     Population count    

HTML PDF Collect

Parallel Authentication Modes Based on Double Blocks or Key Counter

Huang Yuhua,Huai Aiqun,Song Yubo

Strategic Study of CAE 2004, Volume 6, Issue 7,   Pages 70-74

Abstract:

The CBC - MAC mode is not a parallel one. A parallel authentication mode (PKCB) based on double blocks was put forward in this paper. The PKCB mode had a marked improvement on security & speed over parallel authentication mode, PMAC. And it may be combined with the CTR (counter) encryption mode to form a full block cipher mode. On this ground, another parallel authentication mode (KCTR - MAC) based on key counter was advanced. As compared with the PMAC mode, the KCTR - MAC mode had a marked improvement on security, while its speed did not become lower. The KCTR - MAC authentication mode may be combined with the CTR (counter) encryption mode to form a full block cipher mode (2CTR)，too. The 2CTR mode had a performance advantage over the standard mode, CCM (CTR with CBC - MAC). And it was a fast, practicable mode with security.

Keywords： authentication mode     CBC - MAC mode     PMAC mode     CTR mode     CCM mode    

HTML PDF Collect

Title Author Date Type Operation

Forget less, count better: a domain-incremental self-distillation learning benchmark for lifelong crowd counting

Jiaqi GAO, Jingqi LI, Hongming SHAN, Yanyun QU, James Z. WANG, Fei-Yue WANG, Junping ZHANG

Journal Article

A novel convolutional neural network method for crowd counting

Jie-hao Huang, Xiao-guang Di, Jun-de Wu, Ai-yue Chen,18s004055@hit.edu.cn,dixiaoguang@hit.edu.cn

Journal Article

Aggregated context network for crowd counting

Si-yue Yu, Jian Pu,51174500148@stu.ecnu.edu.cn,jianpu@fudan.edu.cn

Journal Article

Crowd modeling based on purposiveness and a destination-driven analysis method

Ning Ding, Weimin Qi, Huihuan Qian,hhqian@cuhk.edu.cn

Journal Article

Research on Social Risk of the Massing Crowd in Public Venues

Li Jianfeng,Liu Mao,Sui Xiaolin

Journal Article

Low-power, high-speed, and area-efficient sequential circuits by quantum-dot cellular automata: T-latch and counter study

Mohammad GHOLAMI, Zaman AMIRZADEH

Journal Article

Controller area network node reliability assessment based on observable node information

Lei-ming ZHANG, Long-hao TANG, Yong LEI

Journal Article

An efficient counter-based Wallace-tree multiplier with a hybrid full adder core for image blending

Ayoub SADEGHI, Nabiollah SHIRI, Mahmood RAFIEE, Mahsa TAHGHIGH

Journal Article

Current Status of Blood Lipid Level and Prevention Strategy of Lipid Disorder in Chinese Population

Chen Haozhu,Jin Xuejuan

Journal Article

Identification of important factors influencing nonlinear counting systems

Xinmin ZHANG, Jingbo WANG, Chihang WEI, Zhihuan SONG,xinminzhang@zju.edu.cn,wangjingbobo@zju.edu.cn,chhwei@zju.edu.cn,songzhihuan@zju.edu.cn

Journal Article

Detecting interaction/complexitywithin crowd movements using braid entropy

Murat AKPULAT, Murat EKİNCİ

Journal Article

MCDB Monte Carlo dosimetry code system and its applications

Deng Li,Li Gang,Chen Chaobin,Ye Tao

Journal Article

Processing and analysis of data from microwave humidity sounder onboard FY-3A satellite

He Jieying,Zhang Shengwei

Journal Article

Binary neural networks for speech recognition

Yan-min QIAN, Xu XIANG

Journal Article

Parallel Authentication Modes Based on Double Blocks or Key Counter

Huang Yuhua,Huai Aiqun,Song Yubo

Journal Article