第 1 条，共 239 条

标题: A level set method with shape priors by using locality preserving projections

作者: Wang, B (Wang, Bin); Gao, XB (Gao, Xinbo); Li, J (Li, Jie); Li, XL (Li, Xuelong); Tao, DC (Tao, Dacheng)

来源出版物: NEUROCOMPUTING  卷: 170  页: 188-200  DOI: 10.1016/j.neucom.2014.07.086  出版年: DEC 25 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 15

使用次数 (2013 年至今): 22

摘要: A novel level set method (LSM) with the constraint of shape priors is proposed to implement a selective image segmentation. Firstly, the shape priors are aligned by using image moment to deprive the spatial related information. Secondly, the aligned shape priors are projected into the subspace expanded by using locality preserving projection to measure the similarity between the shapes. Finally, a new energy functional is built by combing data-driven and shape-driven energy items to implement a selective image segmentation method. We assess the proposed method and some representative LSMs on the synthetic, medical and natural images, the results suggest that the proposed one is superior to the pure data-driven LSMs and the representative LSMs with shape priors. (C) 2015 Elsevier B.V. All rights reserved.

入藏号: WOS:000361256000019

语种: English

文献类型: Article; Proceedings Paper

会议名称: International Conference on Intelligent Science and Intelligent Data Engineering (IScIDE)

会议日期: JUL 31-AUG 02, 2013

会议地点: Beijing, PEOPLES R CHINA

作者关键词: Selective image segmentation; Active contour; Level set method; Shape priors; Locality preserving projections

KeyWords Plus: GEODESIC ACTIVE CONTOURS; IMAGE SEGMENTATION; EVOLUTION; EDGES; FLOW

地址: [Wang, Bin; Gao, Xinbo; Li, Jie] Xidian Univ, Sch Elect Engn, Xian 710071, Shaanxi, Peoples R China. 
[Li, Xuelong] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key La Boratory Transient Opt & Photon, Ctr OPT IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China. 
[Tao, Dacheng] Univ Technol Sydney, Ctr Quantum Computat & Intelligent Syst, Sydney, NSW 2007, Australia.

通讯作者地址: Wang, B (通讯作者),Xidian Univ, Sch Elect Engn, Xian 710071, Shaanxi, Peoples R China.

作者识别号:

出版商: ELSEVIER SCIENCE BV

出版商地址: PO BOX 211, 1000 AE AMSTERDAM, NETHERLANDS

Web of Science 类别: Computer Science, Artificial Intelligence

研究方向: Computer Science

IDS 号: CR3TM

ISSN: 0925-2312

eISSN: 1872-8286

29 字符的来源出版物名称缩写: NEUROCOMPUTING

ISO 来源出版物缩写: Neurocomputing

来源出版物页码计数: 13

第 2 条，共 239 条

标题: Generation and three-dimensional characterization of complex nondiffracting optical beams

作者: Yu, XH (Yu Xiang-Hua); Yao, BL (Yao Bao-Li); Lei, M (Lei Ming); Yan, SH (Yan Shao-Hui); Yang, YL (Yang Yan-Long); Li, RZ (Li Run-Ze); Cai, YN (Cai Ya-Nan)

来源出版物: ACTA PHYSICA SINICA  卷: 64  期: 24  文献号: 244203  DOI: 10.7498/aps.64.244203  出版年: DEC 20 2015  

Web of Science 核心合集中的 "被引频次": 1

被引频次合计: 1

使用次数 (最近 180 天): 5

使用次数 (2013 年至今): 5

摘要: Nondiffracting optical beams play an important role in contemporary optics due to their special propagation characteristics, i.e., nondiffracting in a diffraction-free zone, shape recovering behind obstacles or self-healing property. Liquid crystal spatial light modulators (LC-SLM) are widely used for generating nondiffracting optical beams in virtue of programmable and dynamic features. In this paper, we propose a complex amplitude modulation technique that can encode any scalar complex fields for generating the complex nondiffracting beams. Before experiment, the phase modulation curve of the phase-only LC-SLM is optimized into being linear in a range of 0-2 pi by gamma correction in the way of variable binary phase gratings. Then, we experimentally generate the nonaccelerating beams, e.g., two zero-order Bessel beams with variable intensity distributions, and the nondiffracting petal-like beams generated by interfering with two coaxial Bessel beams. By scanning a reflection mirror near the focal region along the optical axis, a stack of two-dimensional images is acquired, and then a three-dimensional intensity profile of the beam is reconstructed with a software. We also experimentally demonstrate a new kind of multi-main-lobe accelerating beam with parabolic accelerating trajectory by modifying the spatial spectrum of classical Airy beam. Compared with the so-called vectorial accelerating beam with multiple main lobes in spheroidal coordinates, our generated two-main-lobe accelerating beam has a very high energy efficiency. The self-healing property of the two-main-lobe accelerating beam is also demonstrated. The presented technique can generate a variety of complex nondiffracting optical beams rapidly and obtain their three-dimensional intensity distributions accurately, which has potential applications in the fields of optical microscope, optical date storage, optical trapping, optical micromachining, etc.

入藏号: WOS:000368828600024

语种: Chinese

文献类型: Article

作者关键词: modulation of complex amplitude; nondiffracting optical beams; 3D reconstruction of optical beams

KeyWords Plus: SPATIAL LIGHT MODULATORS; AIRY BEAMS; AMPLITUDE; FIELDS; MANIPULATION; TRAPS; LASER

地址: [Yu Xiang-Hua; Yao Bao-Li; Lei Ming; Yan Shao-Hui; Yang Yan-Long; Li Run-Ze; Cai Ya-Nan] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

通讯作者地址: Yao, BL (通讯作者),Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

电子邮件地址: yaobl@opt.ac.cn

出版商: CHINESE PHYSICAL SOC

出版商地址: P O BOX 603, BEIJING 100080, PEOPLES R CHINA

Web of Science 类别: Physics, Multidisciplinary

研究方向: Physics

IDS 号: DB9IC

ISSN: 1000-3290

29 字符的来源出版物名称缩写: ACTA PHYS SIN-CH ED

ISO 来源出版物缩写: Acta Phys. Sin.

来源出版物页码计数: 8

基金资助致谢:

Project supported by the National Basic Research Program of China (Grant No. 2012CB921900) and the National Natural Science Foundation of China (Grant Nos. 81427802, 61275193).

第 3 条，共 239 条

标题: Hydrothermal synthesis of TiO2 nanostructure films and their photoelectrochemical properties

作者: Miao, H (Miao, Hui); Hu, XY (Hu, Xiaoyun); Fan, J (Fan, Jun); Li, CB (Li, Chaoben); Sun, Q (Sun, Qian); Hao, YY (Hao, Yuanyuan); Zhang, GW (Zhang, Guowei); Bai, JT (Bai, Jintao); Hou, X (Hou, Xun)

来源出版物: APPLIED SURFACE SCIENCE  卷: 358  页: 418-424  DOI: 10.1016/j.apsusc.2015.08.212  子辑: A  出版年: DEC 15 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 31

使用次数 (2013 年至今): 31

摘要: In this study, anatase TiO2 nanostructured films were successfully prepared via Ti foil-assisted hydrothermal synthesis method without any surfactants. With the increase of hydrothermal time from 3 to 12 h, the surface morphologies of the films changed from nanotubes to nanoflowers and the thickness of the films increased from 2.5 to 4.2 mu m. The nanoflowers were seated on a layer of oriented nanotubes and consisted of nanosheets. Both nanotubes and nanosheets were multi-walled, comprising interconnected [TiO6] octahedra. Ti foil and acid-treated Ti foil both formed pieces of thin films on the surface with nanotubes and nanoflowers, respectively. Photoelectrochemistry tests showed that the photocurrent response of the nanotube film was nearly 13.8 mu A/cm(2), which was approximately three times higher than that of nanoflowers film. This indicates that the nanotubes more successfully transferred electrons to the surface of the film as a result of nanotubes providing direct pathways for electrons from the point of injection to the Ti foil electrode. The ultraviolet (UV)-visible-near infrared (NIR) absorption spectra indicated that the nanotube film had a narrower band gap as compared to the nanoflower film. In other words, a relatively low hydrothermal treatment temperature or relatively short hydrothermal treatment time is better for preparing excellent photocurrent response films. (C) 2015 Elsevier B.V. All rights reserved.

入藏号: WOS:000366220400052

语种: English

文献类型: Article

作者关键词: TiO2; Nanostructure; Film; Photoelectrochemistry

KeyWords Plus: SENSITIZED SOLAR-CELLS; NANOTUBE ARRAYS; ANATASE TIO2; SURFACE; NANOWIRES; OXIDATION; RUTILE; WATER

地址: [Miao, Hui; Hu, Xiaoyun; Li, Chaoben; Sun, Qian; Hao, Yuanyuan; Zhang, Guowei; Bai, Jintao; Hou, Xun] NW Univ Xian, Sch Phys, Xian 710069, Peoples R China. 
[Miao, Hui; Hu, Xiaoyun; Bai, Jintao; Hou, Xun] NW Univ Xian, Natl Photoelect Technol & Funct Mat & Applicat Sc, Xian 710069, Peoples R China. 
[Fan, Jun] NW Univ Xian, Sch Chem Engn, Xian 710069, Peoples R China. 
[Hou, Xun] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Technol, Xian 710068, Peoples R China.

通讯作者地址: Hu, XY (通讯作者),NW Univ Xian, Sch Phys, Xian 710069, Peoples R China.

电子邮件地址: hxy3275@nwu.edu.cn; fanjun@nwu.edu.cn

出版商: ELSEVIER SCIENCE BV

出版商地址: PO BOX 211, 1000 AE AMSTERDAM, NETHERLANDS

Web of Science 类别: Chemistry, Physical; Materials Science, Coatings & Films; Physics, Applied; Physics, Condensed Matter

研究方向: Chemistry; Materials Science; Physics

IDS 号: CY2EA

ISSN: 0169-4332

eISSN: 1873-5584

29 字符的来源出版物名称缩写: APPL SURF SCI

ISO 来源出版物缩写: Appl. Surf. Sci.

来源出版物页码计数: 7

基金资助致谢:

This work has been supported by the National Natural Science Foundation of China (Grant Nos. 51372201, 21476183), the Shaanxi Provincial Natural Science Foundation (Grant No. 2015JM5159), and the Research Fund for the Doctoral Program of Higher Education (Grant No. 20136101110009).

第 4 条，共 239 条

标题: Fraunhofer diffraction of Laguerre-Gaussian beam caused by a dynamic superposed dual-triangular aperture

作者: Li, XZ (Li, Xinzhong); Tai, YP (Tai, Yuping); Nie, ZG (Nie, Zhaogang); Wang, H (Wang, Hui); Li, HH (Li, Hehe); Wang, JG (Wang, Jingge); Tang, J (Tang, Jie); Wang, YS (Wang, Yishan)

来源出版物: OPTICAL ENGINEERING  卷: 54  期: 12  文献号: 123113  DOI: 10.1117/1.OE.54.12.123113  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 4

使用次数 (2013 年至今): 4

摘要: We investigate the Fraunhofer diffraction of a Laguerre-Gaussian (LG) beam incident on a dynamic superposed dual-triangular aperture. The evolution of the diffraction pattern from this aperture is analyzed experimentally and theoretically. A special aperture, called the hex-star triangular aperture, demonstrates interesting diffraction patterns. Further, the diffraction properties of integer, half-integer, and fractional orders of topological charges at the Fraunhofer zone are studied by using the hex-star triangular aperture. This study can provide additional information to enhance the understanding of the diffraction properties of the LG beam transmitted through a complex aperture. (C) 2015 Society of Photo-Optical Instrumentation Engineers (SPIE)

入藏号: WOS:000368418300016

语种: English

文献类型: Article

作者关键词: diffraction; optical vortices; Laguerre-Gaussian beam; dynamic aperture

KeyWords Plus: ORBITAL ANGULAR-MOMENTUM; OPTICAL VORTEX BEAM; TOPOLOGICAL CHARGE; VORTICES; PROPAGATION; GENERATION; BIRTH

地址: [Li, Xinzhong; Wang, Hui; Li, Hehe; Wang, Jingge] Henan Univ Sci & Technol, Sch Phys & Engn, Luoyang 471003, Peoples R China. 
[Tai, Yuping] Henan Univ Sci & Technol, Sch Chem Engn & Pharmaceut, Luoyang 471003, Peoples R China. 
[Nie, Zhaogang] Nanyang Technol Univ, Sch Phys & Math Sci, Singapore 637371, Singapore. 
[Tang, Jie; Wang, Yishan] Chinese Acad Sci, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

通讯作者地址: Li, XZ (通讯作者),Henan Univ Sci & Technol, Sch Phys & Engn, 263 Kaiyuan Rd, Luoyang 471003, Peoples R China.

电子邮件地址: lixinzhong7922@163.com

出版商: SPIE-SOC PHOTO-OPTICAL INSTRUMENTATION ENGINEERS

出版商地址: 1000 20TH ST, PO BOX 10, BELLINGHAM, WA 98225 USA

Web of Science 类别: Optics

研究方向: Optics

IDS 号: DB3NL

ISSN: 0091-3286

eISSN: 1560-2303

29 字符的来源出版物名称缩写: OPT ENG

ISO 来源出版物缩写: Opt. Eng.

来源出版物页码计数: 10

基金资助致谢:

This work was supported by the National Natural Science Foundation of China (Grant No. 61205086, 11404097, and 11504091), Open Research Fund of State Key Laboratory of Transient Optics and Photonics, Chinese Academy of Sciences (SKLST201203), and Key Research Project of Education Minister of Henan Province (Grant No. 12B140006).

第 5 条，共 239 条

标题: Attitude Disturbance Correction for Binocular Vision Cameras in Rendezvous and Docking

作者: Zhai, Y (Zhai, You); Zeng, L (Zeng, Luan); Xiong, W (Xiong, Wei); Shan, QS (Shan, Qiusha)

来源出版物: IEEE AEROSPACE AND ELECTRONIC SYSTEMS MAGAZINE  卷: 30  期: 12  页: 34-40  DOI: 10.1109/MAES.2015.150019  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 4

使用次数 (2013 年至今): 4

入藏号: WOS:000368406300004

语种: English

文献类型: Article

地址: [Zhai, You; Zeng, Luan; Xiong, Wei] Equipment Acad PLA, Beijing 101416, Peoples R China. 
[Shan, Qiusha] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Xian, Peoples R China.

通讯作者地址: Zhai, Y (通讯作者),Equipment Acad PLA, Key Lab, Bayi Rd 172,Mailbox 3380, Beijing 101416, Peoples R China.

电子邮件地址: youyou1952@sina.com

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Engineering, Aerospace; Engineering, Electrical & Electronic

研究方向: Engineering

IDS 号: DB3JF

ISSN: 0885-8985

eISSN: 1557-959X

29 字符的来源出版物名称缩写: IEEE AERO EL SYS MAG

ISO 来源出版物缩写: IEEE Aerosp. Electron. Syst. Mag.

来源出版物页码计数: 7

第 6 条，共 239 条

标题: Degradation of high power single emitter laser modules using nanosilver paste in continuous pulse conditions

作者: Yan, HD (Yan, Haidong); Mei, YH (Mei, Yunhui); Li, X (Li, Xin); Zhang, P (Zhang, Pu); Lu, GQ (Lu, Guo-Quan)

来源出版物: MICROELECTRONICS RELIABILITY  卷: 55  期: 12  特刊: SI  页: 2532-2541  DOI: 10.1016/j.microrel.2015.07.037  子辑: A  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 5

使用次数 (2013 年至今): 5

摘要: This paper is mainly reported to a pulse reliability investigation of high power single emitter laser modules with nanosilver paste. Comparative experiments in continuous pulse conditions for the laser modules packaged with nanosilver paste, indium and AuSn solders were conducted. The results indicate that the laser modules attached by nanosilver paste have a longer-term lifetime than those with indium and AuSn solders in continuous pulse conditions. Transient thermal behavior and coupled thermo-mechanical behavior in continuous pulse conditions are simulated by finite element method (FEM). A semi-empirical model based on Arrhenius relationship is established to provide relative reliability assessments for laser modules by combining with the simulating results. (C) 2015 Elsevier Ltd. All rights reserved.

入藏号: WOS:000367773000010

语种: English

文献类型: Article

作者关键词: Nanosilver paste; Laser module; Die-attach interface; Pulse mode; Reliability

KeyWords Plus: DIODE BARS; SEMICONDUCTOR-LASERS; THERMAL-PROPERTIES; INTERFACE MATERIAL; INDIUM SOLDER; AU/SN SOLDER; RELIABILITY; TEMPERATURE; PACKAGES; FAILURE

地址: [Yan, Haidong; Mei, Yunhui] Tianjin Univ, Minist Educ, Key Lab Adv Ceram & Machining Technol, Tianjin 300072, Peoples R China. 
[Yan, Haidong; Mei, Yunhui; Li, Xin] Tianjin Univ, Tianjin Key Lab Adv Joining Technol, Tianjin 300072, Peoples R China. 
[Yan, Haidong; Mei, Yunhui; Li, Xin] Tianjin Univ, Sch Mat Sci & Engn, Tianjin 300072, Peoples R China. 
[Zhang, Pu] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China. 
[Lu, Guo-Quan] Virginia Tech, Dept Mat Sci & Engn, Blacksburg, VA 24061 USA.

通讯作者地址: Mei, YH (通讯作者),Tianjin Univ, Minist Educ, Key Lab Adv Ceram & Machining Technol, Tianjin 300072, Peoples R China.

电子邮件地址: yunhui@tju.edu.cn

出版商: PERGAMON-ELSEVIER SCIENCE LTD

出版商地址: THE BOULEVARD, LANGFORD LANE, KIDLINGTON, OXFORD OX5 1GB, ENGLAND

Web of Science 类别: Engineering, Electrical & Electronic; Nanoscience & Nanotechnology; Physics, Applied

研究方向: Engineering; Science & Technology - Other Topics; Physics

IDS 号: DA4MC

ISSN: 0026-2714

29 字符的来源出版物名称缩写: MICROELECTRON RELIAB

ISO 来源出版物缩写: Microelectron. Reliab.

来源出版物页码计数: 10

基金资助致谢:

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 61334010), and the Tianjin Municipal Natural Science Foundation (under grant Nos. 13JCQNJC06600, 13JCZDJC33600, and 13JCQNJC02400). The work was supported by and corresponds to Prof. Yunhui Mei of Tianjin University.

第 7 条，共 239 条

标题: Yb-doped passively mode-locked fiber laser with Bi2Te3-deposited

作者: Li, L (Li Lu); Yan, PG (Yan Pei-Guang); Wang, YG (Wang Yong-Gang); Duan, LN (Duan Li-Na); Sun, H (Sun Hang); Si, JH (Si Jin-Hai)

来源出版物: CHINESE PHYSICS B  卷: 24  期: 12  文献号: 124204  DOI: 10.1088/1674-1056/24/12/124204  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 5

使用次数 (2013 年至今): 5

摘要: In this study we present an all-normal-dispersion Yb-doped fiber laser passively mode-locked with topological insulator (Bi2Te3) saturable absorber. The saturable absorber device is fabricated by depositing Bi2Te3 on a tapered fiber through using pulsed laser deposition (PLD) technology, which can give rise to less non-saturable losses than most of the solution processing methods. Owing to the long interaction length, Bi2Te3 is not exposed to high optical power, which allows the saturable absorber device to work in a high power regime. The modulation depth of this kind of saturable absorber is measured to be 10%. By combining the saturable absorber device with Yb-doped fiber laser, a mode-locked pulse operating at a repetition rate of 19.8 MHz is achieved. The 3-dB spectral width and pulse duration are measured to be 1.245 nm and 317 ps, respectively.

入藏号: WOS:000366977800026

语种: English

文献类型: Article

作者关键词: topological insulator; mode-locked lasers; pulsed laser deposition technology; fiber lasers

KeyWords Plus: EVANESCENT-FIELD INTERACTION; TOPOLOGICAL INSULATOR BI2SE3; MOS2 SATURABLE ABSORBER; SINGLE DIRAC CONE; BI2TE3; GENERATION; SOLITONS; LOCKING; SURFACE; GHZ

地址: [Li Lu; Wang Yong-Gang; Duan Li-Na; Sun Hang] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China. 
[Yan Pei-Guang] Shenzhen Univ, Shenzhen Key Lab Engn, Shenzhen 518060, Peoples R China. 
[Li Lu; Si Jin-Hai] Xi An Jiao Tong Univ, Minist Educ, Key Lab Phys Elect & Devices, Xian 710049, Peoples R China. 
[Li Lu; Si Jin-Hai] Xi An Jiao Tong Univ, Sch Elect & Informat Engn, Shaanxi Key Lab Informat Photon Tech, Xian 710049, Peoples R China. 
[Li Lu; Duan Li-Na] Univ Chinese Acad Sci CAS, Beijing 100049, Peoples R China.

通讯作者地址: Wang, YG (通讯作者),Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

电子邮件地址: chinawygxjw@opt.ac.cn

出版商: IOP PUBLISHING LTD

出版商地址: TEMPLE CIRCUS, TEMPLE WAY, BRISTOL BS1 6BE, ENGLAND

Web of Science 类别: Physics, Multidisciplinary

研究方向: Physics

IDS 号: CZ3AZ

ISSN: 1674-1056

eISSN: 1741-4199

29 字符的来源出版物名称缩写: CHINESE PHYS B

ISO 来源出版物缩写: Chin. Phys. B

来源出版物页码计数: 5

基金资助致谢:

Project supported by the National Natural Science Foundation of China (Grant No. 61378024) and the Natural Science Fund of Guangdong Province, China (Grant No. S2013010012235).

第 8 条，共 239 条

标题: Mining Spatial-Temporal Patterns and Structural Sparsity for Human Motion Data Denoising

作者: Feng, YF (Feng, Yinfu); Ji, MM (Ji, Mingming); Xiao, J (Xiao, Jun); Yang, XS (Yang, Xiaosong); Zhang, JJ (Zhang, Jian J.); Zhuang, YT (Zhuang, Yueting); Li, XL (Li, Xuelong)

来源出版物: IEEE TRANSACTIONS ON CYBERNETICS  卷: 45  期: 12  页: 2693-2706  DOI: 10.1109/TCYB.2014.2381659  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 8

使用次数 (2013 年至今): 8

摘要: Motion capture is an important technique with a wide range of applications in areas such as computer vision, computer animation, film production, and medical rehabilitation. Even with the professional motion capture systems, the acquired raw data mostly contain inevitable noises and outliers. To denoise the data, numerous methods have been developed, while this problem still remains a challenge due to the high complexity of human motion and the diversity of real-life situations. In this paper, we propose a data-driven-based robust human motion denoising approach by mining the spatial-temporal patterns and the structural sparsity embedded in motion data. We first replace the regularly used entire pose model with a much fine-grained partlet model as feature representation to exploit the abundant local body part posture and movement similarities. Then, a robust dictionary learning algorithm is proposed to learn multiple compact and representative motion dictionaries from the training data in parallel. Finally, we reformulate the human motion denoising problem as a robust structured sparse coding problem in which both the noise distribution information and the temporal smoothness property of human motion have been jointly taken into account. Compared with several state-of-the-art motion denoising methods on both the synthetic and real noisy motion data, our method consistently yields better performance than its counterparts. The outputs of our approach are much more stable than that of the others. In addition, it is much easier to setup the training dataset of our method than that of the other data-driven-based methods.

入藏号: WOS:000365320300006

PubMed ID: 25561602

语种: English

文献类型: Article

作者关键词: Human motion denoising; l(2,p)-norm; Microsoft Kinect; motion capture data; robust dictionary learning; robust structured sparse coding

KeyWords Plus: ACTION RECOGNITION; CAPTURE; REPRESENTATION; IMAGES; ANIMATION; WAVELETS; PURSUIT; NOISE

地址: [Feng, Yinfu; Ji, Mingming; Xiao, Jun; Zhuang, Yueting] Zhejiang Univ, Coll Comp Sci, Hangzhou 310027, Zhejiang, Peoples R China. 
[Yang, Xiaosong; Zhang, Jian J.] Bournemouth Univ, Natl Ctr Comp Animat, Poole BH12 5BB, Dorset, England. 
[Li, Xuelong] Chinese Acad Sci, Ctr Opt Imagery Anal & Learning, State Key Lab Transient Opt & Photon, Xian Inst Opt & Precis Mech, Xian 710119, Peoples R China.

通讯作者地址: Xiao, J (通讯作者),Zhejiang Univ, Coll Comp Sci, Hangzhou 310027, Zhejiang, Peoples R China.

电子邮件地址: junx@cs.zju.edu.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Computer Science, Artificial Intelligence; Computer Science, Cybernetics

研究方向: Computer Science

IDS 号: CW9LN

ISSN: 2168-2267

eISSN: 2168-2275

29 字符的来源出版物名称缩写: IEEE T CYBERNETICS

ISO 来源出版物缩写: IEEE T. Cybern.

来源出版物页码计数: 14

基金资助致谢:

This work was supported in part by the National Key Basic Research Program of China under Grant 2012CB316400, in part by the National High Technology Research and Development Program under Grant 2012AA011502, in part by the Zhejiang Provincial Natural Science Foundation of China under Grant LY13F020001, in part by the Fundamental Research Funds for the Central Universities under Grant 2014FZA5013, in part by Zhejiang Province Public Technology Applied Research Projects under Grant 2014C33090, and in part by the grant of the Sino-U.K. Higher Education Research Partnership for Ph.D. student's project funded by the Department of Business, Innovation, and Skills of the British Government and the Ministry of Education of China. This paper was recommended by Associate Editor L. Shao.

第 9 条，共 239 条

标题: Efficient Video Stitching Based on Fast Structure Deformation

作者: Li, J (Li, Jing); Xu, W (Xu, Wei); Zhang, JG (Zhang, Jianguo); Zhang, MJ (Zhang, Maojun); Wang, ZM (Wang, Zhengming); Li, XL (Li, Xuelong)

来源出版物: IEEE TRANSACTIONS ON CYBERNETICS  卷: 45  期: 12  页: 2707-2719  DOI: 10.1109/TCYB.2014.2381774  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 8

使用次数 (2013 年至今): 8

摘要: In computer vision, video stitching is a very challenging problem. In this paper, we proposed an efficient and effective wide-view video stitching method based on fast structure deformation that is capable of simultaneously achieving quality stitching and computational efficiency. For a group of synchronized frames, firstly, an effective double-seam selection scheme is designed to search two distinct but structurally corresponding seams in the two original images. The seam location of the previous frame is further considered to preserve the inter-frame consistency. Secondly, along the double seams, 1-D feature detection and matching is performed to capture the structural relationship between the two adjacent views. Thirdly, after feature matching, we propose an efficient algorithm to linearly propagate the deformation vectors to eliminate structure misalignment. At last, image intensity misalignment is corrected by rapid gradient fusion based on the successive over relaxation iteration (SORI) solver. A principled solution to the initialization of the SORI significantly reduced the number of iterations required. We have compared favorably our method with seven state-of-the-art image and video stitching algorithms as well as traditional ones. Experimental results show that our method outperforms the existing ones compared in terms of overall stitching quality and computational efficiency.

入藏号: WOS:000365320300007

PubMed ID: 25561603

语种: English

文献类型: Article

作者关键词: Computational efficiency; computer vision; intensity misalignment; machine learning; structure misalignment; video stitching

KeyWords Plus: TEXTURE SYNTHESIS; IMAGE MOSAICS; REGISTRATION; DOMAIN; RECOGNITION; FEATURES; RANGE

地址: [Li, Jing; Xu, Wei; Zhang, Maojun] Natl Univ Def Technol, Coll Informat Syst & Management, Changsha 410073, Hunan, Peoples R China. 
[Zhang, Jianguo] Univ Dundee, Sch Comp, Dundee DD1 4HN, Scotland. 
[Wang, Zhengming] Natl Univ Def Technol, Coll Sci, Changsha 410073, Hunan, Peoples R China. 
[Li, Xuelong] Chinese Acad Sci, Ctr Opt IMagery Anal & Learning OPTIMAL, State Key Lab Transient Opt & Photon, Xian Inst Opt & Precis Mech, Xian 710119, Shaanxi, Peoples R China.

通讯作者地址: Li, J (通讯作者),Natl Univ Def Technol, Coll Informat Syst & Management, Changsha 410073, Hunan, Peoples R China.

电子邮件地址: jingli@nudt.edu.cn; weixu@nudt.edu.cn; jgzhang@computing.dundee.ac.uk; xuelong_li@opt.ac.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Computer Science, Artificial Intelligence; Computer Science, Cybernetics

研究方向: Computer Science

IDS 号: CW9LN

ISSN: 2168-2267

eISSN: 2168-2275

29 字符的来源出版物名称缩写: IEEE T CYBERNETICS

ISO 来源出版物缩写: IEEE T. Cybern.

来源出版物页码计数: 13

基金资助致谢:

This work was supported in part by the National Natural Science Foundation of China under Grant 61403403 and Grant 61125106, in part by the National University of Defense and Technology under Project JC13-05-02, and in part by the Shaanxi Key Innovation Team of Science and Technology under Grant 2012KCT-04. This paper was recommended by Associate Editor L. Shao.

第 10 条，共 239 条

标题: Learning a Tracking and Estimation Integrated Graphical Model for Human Pose Tracking

作者: Zhao, L (Zhao, Lin); Gao, XB (Gao, Xinbo); Tao, DC (Tao, Dacheng); Li, XL (Li, Xuelong)

来源出版物: IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS  卷: 26  期: 12  页: 3176-3186  DOI: 10.1109/TNNLS.2015.2411287  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 2

被引频次合计: 2

使用次数 (最近 180 天): 0

使用次数 (2013 年至今): 0

摘要: We investigate the tracking of 2-D human poses in a video stream to determine the spatial configuration of body parts in each frame, but this is not a trivial task because people may wear different kinds of clothing and may move very quickly and unpredictably. The technology of pose estimation is typically applied, but it ignores the temporal context and cannot provide smooth, reliable tracking results. Therefore, we develop a tracking and estimation integrated model (TEIM) to fully exploit temporal information by integrating pose estimation with visual tracking. However, joint parsing of multiple articulated parts over time is difficult, because a full model with edges capturing all pairwise relationships within and between frames is loopy and intractable. In previous models, approximate inference was usually resorted to, but it cannot promise good results and the computational cost is large. We overcome these problems by exploring the idea of divide and conquer, which decomposes the full model into two much simpler tractable submodels. In addition, a novel two-step iteration strategy is proposed to efficiently conquer the joint parsing problem. Algorithmically, we design TEIM very carefully so that: 1) it enables pose estimation and visual tracking to compensate for each other to achieve desirable tracking results; 2) it is able to deal with the problem of tracking loss; and 3) it only needs past information and is capable of tracking online. Experiments are conducted on two public data sets in the wild with ground truth layout annotations, and the experimental results indicate the effectiveness of the proposed TEIM framework.

入藏号: WOS:000365312800017

PubMed ID: 25826809

语种: English

文献类型: Article

作者关键词: Pictorial structure (PS); pose estimation; pose tracking; visual tracking

KeyWords Plus: DISCRIMINANT-ANALYSIS; VISUAL TRACKING; PICTORIAL STRUCTURES; FLEXIBLE MIXTURES; RECOGNITION; APPEARANCE; RETRIEVAL; PARTS

地址: [Zhao, Lin; Gao, Xinbo] Xidian Univ, Sch Elect Engn, Video & Image Proc Syst VIPS Lab, Xian 710071, Peoples R China. 
[Tao, Dacheng] Univ Technol Sydney, Ctr Quantum Computat & Intelligent Syst, Ultimo, NSW 2007, Australia. 
[Tao, Dacheng] Univ Technol Sydney, Fac Engn & Informat Technol, Ultimo, NSW 2007, Australia. 
[Li, Xuelong] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Ctr OpT IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China.

通讯作者地址: Zhao, L (通讯作者),Xidian Univ, VIPS Lab, Sch Elect Engn, Pattern Recognit & Intelligent Syst, Xian 710071, Peoples R China.

电子邮件地址: sharling1988@gmail.com; xbgao@mail.xidian.edu.cn; dacheng.tao@uts.edu.au; xuelong_li@opt.ac.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Computer Science, Artificial Intelligence; Computer Science, Hardware & Architecture; Computer Science, Theory & Methods; Engineering, Electrical & Electronic

研究方向: Computer Science; Engineering

IDS 号: CW9IQ

ISSN: 2162-237X

eISSN: 2162-2388

29 字符的来源出版物名称缩写: IEEE T NEUR NET LEAR

ISO 来源出版物缩写: IEEE Trans. Neural Netw. Learn. Syst.

来源出版物页码计数: 11

基金资助致谢:

This work is supported by the National Natural Science Foundation of China (Grant Nos. 61125204, 61432014, 61172146, and 61125106), the Fundamental Research Funds for the Central Universities (Grant Nos. BDZ021403 and JB149901), the Program for Changjiang Scholars and Innovative Research Team in University of China (No. IRT13088), the Shaanxi Innovative Research Team for Key Science and Technology (No. 2012KCT-02), Australian Research Council Projects FT-130101457, DP-140102164 and DP-120103730, and by the Key Research Program of the Chinese Academy of Sciences (Grant No. KGZD-EW-T03).

第 11 条，共 239 条

标题: Passively Q-switched Nd:YAG laser with a MoS2 solution saturable absorber

作者: Lin, T (Lin, Tao); Sun, H (Sun, Hang); Wang, X (Wang, Xi); Mao, D (Mao, Dong); Wang, YG (Wang, Yonggang); Li, L (Li, Lu); Duan, LN (Duan, Lina)

来源出版物: LASER PHYSICS  卷: 25  期: 12  文献号: 125805  DOI: 10.1088/1054-660X/25/2/125805  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 4

使用次数 (2013 年至今): 4

摘要: This study presents a passively Q-switched Nd:YAG laser using a MoS2 solution as a saturable absorber. The prepared MoS2 solution saturable absorber features adjustable optical absorption, high heat dissipation, and non-contact damage. By inserting the MoS2 solution saturable absorber into a Z-type Nd:YAG laser cavity, a stable Q-switched laser with the maximum repetition rate of 20.8 KHz centered at 1064.24 nm was achieved, with the shortest pulse width at 2.8 mu s. The results experimentally confirm the potential application of a MoS2 solution saturable absorber for high performance Q-switched solid-state lasers.

入藏号: WOS:000365351200020

语种: English

文献类型: Article

作者关键词: Q-switched; solid-state lasers; MoS2; saturable absorber

KeyWords Plus: ABSORPTION

地址: [Lin, Tao; Sun, Hang] Xian Univ Technol, Dept Elect Engn, Xian 710048, Peoples R China. 
[Wang, Xi; Wang, Yonggang; Li, Lu; Duan, Lina] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China. 
[Mao, Dong] Northwestern Polytech Univ, Minist Educ, Key Lab Space Appl Phys & Chem, Xian 710072, Peoples R China. 
[Mao, Dong] Northwestern Polytech Univ, Sch Sci, Shaanxi Key Lab Opt Informat Technol, Xian 710072, Peoples R China.

通讯作者地址: Lin, T (通讯作者),Xian Univ Technol, Dept Elect Engn, Xian 710048, Peoples R China.

电子邮件地址: lintao@xuat.edu.cn

出版商: IOP PUBLISHING LTD

出版商地址: TEMPLE CIRCUS, TEMPLE WAY, BRISTOL BS1 6BE, ENGLAND

Web of Science 类别: Optics; Physics, Applied

研究方向: Optics; Physics

IDS 号: CW9XJ

ISSN: 1054-660X

eISSN: 1555-6611

29 字符的来源出版物名称缩写: LASER PHYS

ISO 来源出版物缩写: Laser Phys.

来源出版物页码计数: 5

基金资助致谢:

This work was supported by the national natural science foundation of China under grants 61378024, and the science and the open research fund of state key laboratory of transient optics and photonics, Chinese academy of sciences under grants SKLST201304.

第 12 条，共 239 条

标题: Highly efficient nonlinear ellipse rotation in bulk SrTiO3 crystal and its applications

作者: Sun, QB (Sun, Qibing); Liu, HJ (Liu, Hongjun); Huang, N (Huang, Nan); Wang, ZL (Wang, Zhaolu); Li, SP (Li, Shaopeng); Han, J (Han, Jing)

来源出版物: LASER PHYSICS  卷: 25  期: 12  文献号: 125403  DOI: 10.1088/1054-660X/25/12/125403  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 5

使用次数 (2013 年至今): 5

摘要: We investigated the highly efficient, nonlinear ellipse rotation (NER) effect in bulk SrTiO3 crystal, which provides practical application as a pulse cleaner for ultrashort pulses. A highly stable, ultrafast NER effect was realized by using a bulk SrTiO3 crystal with a fast response time, strong optical Kerr effect, high damage threshold, and wide transmission wavelength range. Transmissivity higher than 30% was obtained at a level of several mJ, in which the temporal contrast can be enhanced by about three orders of magnitude. This provides a simple and feasible technology for improving the temporal contrast for the ultrahigh power laser system.

入藏号: WOS:000365351200012

语种: English

文献类型: Article

作者关键词: nonlinear ellipse rotation; SrTiO3 crystal; pulse cleaner; temporal contrast

KeyWords Plus: BIREFRINGENT FIBERS; PULSES; LASER; CLEANER

地址: [Sun, Qibing; Liu, Hongjun; Huang, Nan; Wang, Zhaolu; Li, Shaopeng; Han, Jing] Chinese Acad Sci, State Key Lab Transient Opt & Photon, Xian Inst Opt & Precis Mech, Xian 710119, Peoples R China.

通讯作者地址: Sun, QB (通讯作者),Chinese Acad Sci, State Key Lab Transient Opt & Photon, Xian Inst Opt & Precis Mech, Xian 710119, Peoples R China.

电子邮件地址: liuhongjun@opt.ac.cn

作者识别号:

出版商: IOP PUBLISHING LTD

出版商地址: TEMPLE CIRCUS, TEMPLE WAY, BRISTOL BS1 6BE, ENGLAND

Web of Science 类别: Optics; Physics, Applied

研究方向: Optics; Physics

IDS 号: CW9XJ

ISSN: 1054-660X

eISSN: 1555-6611

29 字符的来源出版物名称缩写: LASER PHYS

ISO 来源出版物缩写: Laser Phys.

来源出版物页码计数: 4

基金资助致谢:

This research is supported by the National Nature Science Foundation of China under Grant 61178023 and 61275134.

第 13 条，共 239 条

标题: Linkage of Dual Frequency Combs Based on All Polarization-Maintaining Femtosecond Fiber Lasers

作者: Xu, X (Xu, Xin); Feng, Y (Feng, Ye); Liu, YS (Liu, Yuanshan); Hu, XH (Hu, Xiaohong); Li, FT (Li, Feitao); Zhang, T (Zhang, Ting); Zhang, W (Zhang, Wei); Yang, Z (Yang, Zhi); Cheng, Z (Cheng, Zhao); Zhao, W (Zhao, Wei); Wang, YS (Wang, Yishan)

来源出版物: IEEE PHOTONICS TECHNOLOGY LETTERS  卷: 27  期: 23  页: 2421-2424  DOI: 10.1109/LPT.2015.2462080  出版年: DEC 1 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 16

使用次数 (2013 年至今): 16

摘要: Perfect linkage of dual frequency combs based on all polarization-maintaining femtosecond fiber lasers is demonstrated. By employing two independent tracking feedback controlling circuits, the repetition frequency (f(r)) and the carrier-envelope offset frequency (f(ceo)) of the slave frequency comb track the master's f(r) and f(ceo), respectively. The correlation coefficients of the dual f(r)'s and the dual f(ceo)'s are up to 0.99 and 0.98, which indicate an ideal linkage between the dual frequency combs. To the best of our knowledge, this is the first report about linkage of dual frequency combs.

入藏号: WOS:000363209500001

语种: English

文献类型: Article

作者关键词: Ultrafast optics; supercontinuum generation; fiber laser

地址: [Xu, Xin; Feng, Ye; Liu, Yuanshan; Hu, Xiaohong; Li, Feitao; Zhang, Ting; Zhang, Wei; Yang, Zhi; Zhao, Wei; Wang, Yishan] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China. 
[Cheng, Zhao] FEMTOLASERS Prod Gmbh, A-1100 Vienna, Austria.

通讯作者地址: Liu, YS (通讯作者),Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

电子邮件地址: xuxin_1007@163.com; ye.feng.tut@gmail.com; liuyuanshan@opt.ac.cn; xhhu@opt.ac.cn; lifeitao@opt.cn; zhangting712@opt.ac.cn; zhangwei@opt.ac.cn; yanghzhi@opt.ac.cn; zhao.cheng@femtolasers.com; weiz@opt.ac.cn; yishanwang@opt.ac.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Engineering, Electrical & Electronic; Optics; Physics, Applied

研究方向: Engineering; Optics; Physics

IDS 号: CU0LZ

ISSN: 1041-1135

eISSN: 1941-0174

29 字符的来源出版物名称缩写: IEEE PHOTONIC TECH L

ISO 来源出版物缩写: IEEE Photonics Technol. Lett.

来源出版物页码计数: 4

基金资助致谢:

This work was supported in part by the Natural Science Foundation of China under Grant 61275164, in part by the Chinese Academy of Sciences/State Administration of Foreign Experts Affairs International Partnership Program for Creative Research Teams, and in part by the National Major Scientific Instrumentation Development Program of China under Grant 2011YQ120022. (Xin Xu and Ye Feng contributed equally to this work.) (Corresponding authors: Yuanshan Liu and Yishan Wang.)

第 14 条，共 239 条

标题: Plasmon-induced transparency in terahertz planar metamaterials

作者: Chen, X (Chen, Xu); Fan, WH (Fan, Wen-Hui)

来源出版物: OPTICS COMMUNICATIONS  卷: 356  页: 84-89  DOI: 10.1016/j.optcom.2015.07.063  出版年: DEC 1 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 61

使用次数 (2013 年至今): 64

摘要: A planar metamaterial structure with plasmon-induced transparency effect in terahertz region is proposed and systematic numerical study is presented in this paper. The metamaterial structure is comprised of two different spilt-ring resonators in the same plane. With the destructive interference coupling between these two split-ring resonators, the proposed metamaterial structure exhibits a large transparency window within a broad absorption spectra. Moreover, the origin of transparency window with extremely low absorption and strong dispersion is clarified by two coupled Lorentzian resonators analytical model, and verified by accurate simulation of the electromagnetic wave propagation in the metamaterial structure. The proposed metamaterial opens up the avenue to design micro-sized functional devices used in switching, modulation, and slowing down terahertz waves. (C) 2015 Elsevier B.V. All rights reserved.

入藏号: WOS:000362603600016

语种: English

文献类型: Article

作者关键词: Metamaterials; Far infrared or terahertz; Surface plasmons

KeyWords Plus: ELECTROMAGNETICALLY INDUCED TRANSPARENCY; FANO RESONANCES; MAGNETIC-RESONANCE; ANALOG; FREQUENCIES; LIGHT

地址: [Chen, Xu; Fan, Wen-Hui] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China. 
[Chen, Xu] Univ Chinese Acad Sci, Beijing 100049, Peoples R China.

通讯作者地址: Fan, WH (通讯作者),Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

电子邮件地址: fanwh@opt.ac.cn

出版商: ELSEVIER SCIENCE BV

出版商地址: PO BOX 211, 1000 AE AMSTERDAM, NETHERLANDS

Web of Science 类别: Optics

研究方向: Optics

IDS 号: CT2AE

ISSN: 0030-4018

eISSN: 1873-0310

29 字符的来源出版物名称缩写: OPT COMMUN

ISO 来源出版物缩写: Opt. Commun.

来源出版物页码计数: 6

基金资助致谢:

This work was financially supported by the "Hundreds of Talents Programs" of the Chinese Academy of Sciences (Grant no. J08-029), the Innovative Project of the Chinese Academy of Sciences (Grant no. YYYJ-1123-4), and the CAS/SAFEA International Partnership Program for Creative Research Teams.

第 15 条，共 239 条

标题: Tracking Human Pose Using Max-Margin Markov Models

作者: Zhao, L (Zhao, Lin); Gao, XB (Gao, Xinbo); Tao, DC (Tao, Dacheng); Li, XL (Li, Xuelong)

来源出版物: IEEE TRANSACTIONS ON IMAGE PROCESSING  卷: 24  期: 12  页: 5274-5287  DOI: 10.1109/TIP.2015.2473662  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 7

使用次数 (2013 年至今): 8

摘要: We present a new method for tracking human pose by employing max-margin Markov models. Representing a human body by part-based models, such as pictorial structure, the problem of pose tracking can be modeled by a discrete Markov random field. Considering max-margin Markov networks provide an efficient way to deal with both structured data and strong generalization guarantees, it is thus natural to learn the model parameters using the max-margin technique. Since tracking human pose needs to couple limbs in adjacent frames, the model will introduce loops and will be intractable for learning and inference. Previous work has resorted to pose estimation methods, which discard temporal information by parsing frames individually. Alternatively, approximate inference strategies have been used, which can overfit to statistics of a particular data set. Thus, the performance and generalization of these methods are limited. In this paper, we approximate the full model by introducing an ensemble of two tree-structured sub-models, Markov networks for spatial parsing and Markov chains for temporal parsing. Both models can be trained jointly using the max-margin technique, and an iterative parsing process is proposed to achieve the ensemble inference. We apply our model on three challengeable data sets, which contains highly varied and articulated poses. Comprehensive experimental results demonstrate the superior performance of our method over the state-of-the-art approaches.

入藏号: WOS:000362488900013

语种: English

文献类型: Article

作者关键词: Pose tracking; pose estimation; max-margin; articulated shapes

KeyWords Plus: ACTION RECOGNITION; PICTORIAL STRUCTURES; FLEXIBLE MIXTURES; PEOPLE; VIDEO; FLOW; PROPAGATION; PARTS

地址: [Zhao, Lin] Xidian Univ, Sch Elect Engn, Xian 710071, Peoples R China. 
[Gao, Xinbo] Xidian Univ, Sch Elect Engn, State Key Lab Integrated Serv Networks, Xian 710071, Peoples R China. 
[Tao, Dacheng] Univ Technol, Ctr Quantum Computat & Intelligent Syst, Sydney, NSW 2007, Australia. 
[Tao, Dacheng] Univ Technol, Fac Engn & Informat Technol, Sydney, NSW 2007, Australia. 
[Li, Xuelong] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Ctr Opt IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China.

通讯作者地址: Zhao, L (通讯作者),Xidian Univ, Sch Elect Engn, Xian 710071, Peoples R China.

电子邮件地址: sharling1988@gmail.com; xbgao@mail.xidian.edu.cn; dacheng.tao@uts.edu.au; xuelong_li@opt.ac.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Computer Science, Artificial Intelligence; Engineering, Electrical & Electronic

研究方向: Computer Science; Engineering

IDS 号: CT0MC

ISSN: 1057-7149

eISSN: 1941-0042

29 字符的来源出版物名称缩写: IEEE T IMAGE PROCESS

ISO 来源出版物缩写: IEEE Trans. Image Process.

来源出版物页码计数: 14

基金资助致谢:

This work was supported in part by the National Basic Research Program of China (973 Program) under Grant 2012CB316400, in part by the National Natural Science Foundation of China under Grant 61125204, and Grant 61432014, in part by the Fundamental Research Funds through the Central Universities under Grant BDZ021403 and Grant JB149901, in part by the Program for Changjiang Scholars and Innovative Research Team in University of China under Grant IRT13088, in part by the Shaanxi Innovative Research Team for Key Science and Technology under Grant 2012KCT-02, in part by the Australian Research Council under Project DP-120103730 and Project FT-130101457, and in part by the Key Research Program through the Chinese Academy of Sciences under Grant KGZDEW-T03. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Dimitrios Tzovaras.

第 16 条，共 239 条

标题: A Class of Manifold Regularized Multiplicative Update Algorithms for Image Clustering

作者: Yang, SM (Yang, Shangming); Yi, Z (Yi, Zhang); He, XF (He, Xiaofei); Li, XL (Li, Xuelong)

来源出版物: IEEE TRANSACTIONS ON IMAGE PROCESSING  卷: 24  期: 12  页: 5302-5314  DOI: 10.1109/TIP.2015.2457033  出版年: DEC 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 18

使用次数 (2013 年至今): 19

摘要: Multiplicative update algorithms are important tools for information retrieval, image processing, and pattern recognition. However, when the graph regularization is added to the cost function, different classes of sample data may be mapped to the same subspace, which leads to the increase of data clustering error rate. In this paper, an improved nonnegative matrix factorization (NMF) cost function is introduced. Based on the cost function, a class of novel graph regularized NMF algorithms is developed, which results in a class of extended multiplicative update algorithms with manifold structure regularization. Analysis shows that in the learning, the proposed algorithms can efficiently minimize the rank of the data representation matrix. Theoretical results presented in this paper are confirmed by simulations. For different initializations and data sets, variation curves of cost functions and decomposition data are presented to show the convergence features of the proposed update rules. Basis images, reconstructed images, and clustering results are utilized to present the efficiency of the new algorithms. Last, the clustering accuracies of different algorithms are also investigated, which shows that the proposed algorithms can achieve state-of-the-art performance in applications of image clustering.

入藏号: WOS:000362488900015

PubMed ID: 26186793

语种: English

文献类型: Article

作者关键词: Structure retrieving; convergence analysis; manifold structure; multiplicative algorithms; low rank representation

KeyWords Plus: NONNEGATIVE MATRIX FACTORIZATION; NONLINEAR DIMENSIONALITY REDUCTION; GEOMETRIC FRAMEWORK; SEGMENTATION

地址: [Yang, Shangming] Univ Elect Sci & Technol China, Sch Informat & Software Engn, Chengdu 611731, Peoples R China. 
[Yi, Zhang] Sichuan Univ, Coll Comp Sci, Chengdu 610065, Peoples R China. 
[He, Xiaofei] Zhejiang Univ, State Key Lab CAD & CG, Hangzhou 310058, Zhejiang, Peoples R China. 
[Li, Xuelong] Chinese Acad Sci, Xian Inst Opt & Precis Mech, Ctr Opt IMagery Anal & Learning, State Key Lab Transient Opt & Photon, Xian 710119, Peoples R China.

通讯作者地址: Yang, SM (通讯作者),Univ Elect Sci & Technol China, Sch Informat & Software Engn, Chengdu 611731, Peoples R China.

电子邮件地址: minn003@163.com; zhangyi@scu.edu.cn; xiaofeihe@cad.zju.edu.cn; xuelong_li@nwpu.edu.cn

出版商: IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

出版商地址: 445 HOES LANE, PISCATAWAY, NJ 08855-4141 USA

Web of Science 类别: Computer Science, Artificial Intelligence; Engineering, Electrical & Electronic

研究方向: Computer Science; Engineering

IDS 号: CT0MC

ISSN: 1057-7149

eISSN: 1941-0042

29 字符的来源出版物名称缩写: IEEE T IMAGE PROCESS

ISO 来源出版物缩写: IEEE Trans. Image Process.

来源出版物页码计数: 13

基金资助致谢:

This work was supported in part by the National Science Foundation of China under Grant 61432012, in part by the Shaanxi Key Innovation Team of Science and Technology under Grant 2012KCT-04, and in part by the Program for New Century Excellent Talents in University under Grant NCET-11-0355. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Javier Mateos.

第 17 条，共 239 条

标题: Video-based road detection via online structural learning

作者: Yuan, Y (Yuan, Yuan); Jiang, ZY (Jiang, Zhiyu); Wang, Q (Wang, Qi)

来源出版物: NEUROCOMPUTING  卷: 168  页: 336-347  DOI: 10.1016/j.neucom.2015.05.092  出版年: NOV 30 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 8

使用次数 (2013 年至今): 38

摘要: Video-based road detection is a crucial enabler for the successful development of driver assistant and robot navigation systems. But reliable detection is still on its infancy and deserves further research. In order to adapt to the situation consisting of environmental varieties, an online framework is proposed focusing on exploring the structure cue of the feature vectors. Through the structural support vector machine, the road boundary and non-boundary instances are firstly discriminated. Then they are utilized to fit a complete road boundary. After that, the road region is accordingly inferred and the obtained results are treated as ground truth to update the learned model. Three contributions are claimed in this work: online-learning updating, structural information consideration, and targeted sampling selection. The proposed method is finally evaluated on several challenging videos captured by ourselves. Qualitative and quantitative results show that it outperforms the other competitors. (C) 2015 Elsevier B.V. All rights reserved.

入藏号: WOS:000359165000034

语种: English

文献类型: Article

作者关键词: Computer vision; Machine learning; Road detection; Structural SVM; Online updating; Road boundary

KeyWords Plus: TRAFFIC SCENES; PEDESTRIAN DETECTION; VISION; TRACKING; CLASSIFICATION; SCENARIOS; FEATURES; SALIENCY; SYSTEMS; SCALE

地址: [Yuan, Yuan; Jiang, Zhiyu] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Ctr OPT IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China. 
[Jiang, Zhiyu] Univ Chinese Acad Sci, Beijing 100049, Peoples R China. 
[Wang, Qi] Northwestern Polytech Univ, Sch Comp Sci, Xian 710072, Shaanxi, Peoples R China. 
[Wang, Qi] Northwestern Polytech Univ, Ctr OPT IMagery Anal & Learning OPTIMAL, Xian 710072, Shaanxi, Peoples R China.

通讯作者地址: Wang, Q (通讯作者),Northwestern Polytech Univ, Sch Comp Sci, Xian 710072, Shaanxi, Peoples R China.

电子邮件地址: crabwq@nwpu.edu.cn

出版商: ELSEVIER SCIENCE BV

出版商地址: PO BOX 211, 1000 AE AMSTERDAM, NETHERLANDS

Web of Science 类别: Computer Science, Artificial Intelligence

研究方向: Computer Science

IDS 号: CO4XU

ISSN: 0925-2312

eISSN: 1872-8286

29 字符的来源出版物名称缩写: NEUROCOMPUTING

ISO 来源出版物缩写: Neurocomputing

来源出版物页码计数: 12

基金资助致谢:

This work is supported by the National Basic Research Program of China (Youth 973 Program) (Grant no. 2013CB336500), State Key Program of National Natural Science of China (Grant no. 61232010), National Natural Science Foundation of China (Grant nos. 61172143, 61379094, and 61105012), Natural Science Foundation Research Project of Shaanxi Province (Grant no. 2015JM6264), Fundamental Research Funds for the Central Universities (Grant no. 3102014JC02020G07), and Open Research Fund of Key Laboratory of Spectral Imaging Technology, Chinese Academy of Sciences.

第 18 条，共 239 条

标题: Sparse kernel entropy component analysis for dimensionality reduction of biomedical data

作者: Shi, J (Shi, Jun); Jiang, QK (Jiang, Qikun); Zhang, Q (Zhang, Qi); Huang, QH (Huang, Qinghua); Li, XL (Li, Xuelong)

来源出版物: NEUROCOMPUTING  卷: 168  页: 930-940  DOI: 10.1016/j.neucom.2015.05.032  出版年: NOV 30 2015  

Web of Science 核心合集中的 "被引频次": 0

被引频次合计: 0

使用次数 (最近 180 天): 10

使用次数 (2013 年至今): 17

摘要: Dimensionality reduction is ubiquitous in biomedical applications. A newly proposed spectral dimensionality reduction method, named kernel entropy component analysis (KECA), can reveal the structure related to Renyi entropy of an input space data set. However, each principal component in the Hilbert space depends on all training samples in KECA, causing degraded performance. To overcome this drawback, a sparse KECA (SKECA) algorithm based on a recursive divide-and-conquer (DC) method is proposed in this work. The original large and complex problem of KECA is decomposed into a series of small and simple sub-problems, and then they are solved recursively. The performance of SKECA is evaluated on four biomedical datasets, and compared with KECA, principal component analysis (PCA), kernel PCA (KPCA), sparse PCA and sparse KPCA. Experimental results indicate that the SKECA outperforms conventional dimensionality reduction algorithms, even for high order dimensional features. It suggests that SKECA is potentially applicable to biomedical data processing. (C) 2015 Elsevier B.V. All rights reserved.

入藏号: WOS:000359165000091

语种: English

文献类型: Article

作者关键词: Sparse kernel entropy component analysis; Divide-and-conquer method; Dimensionality reduction; Biomedical data

KeyWords Plus: IMAGE; CLASSIFICATION

地址: [Shi, Jun; Jiang, Qikun; Zhang, Qi] Shanghai Univ, Sch Commun & Informat Engn, Shanghai, Peoples R China. 
[Huang, Qinghua] S China Univ Technol, Sch Elect & Informat Engn, Guangzhou 510641, Guangdong, Peoples R China. 
[Li, Xuelong] Chinese Acad Sci, Xian Inst Opt & Precis Mech, State Key Lab Transient Opt & Photon, Ctr OPT IMagery Anal & Learning OPTIMAL, Xian 710119, Shaanxi, Peoples R China.

通讯作者地址: Huang, QH (通讯作者),S China Univ Technol, Sch Elect & Informat Engn, Guangzhou 510641, Guangdong, Peoples R China.

电子邮件地址: qhhuang@scut.edu.cn

出版商: ELSEVIER SCIENCE BV

出版商地址: PO BOX 211, 1000 AE AMSTERDAM, NETHERLANDS

Web of Science 类别: Computer Science, Artificial Intelligence

研究方向: Computer Science

IDS 号: CO4XU

ISSN: 0925-2312

eISSN: 1872-8286

29 字符的来源出版物名称缩写: NEUROCOMPUTING

ISO 来源出版物缩写: Neurocomputing

来源出版物页码计数: 11

基金资助致谢: