Publications-Theses

題名 應用共變異矩陣描述子及半監督式學習於行人偵測
Semi-supervised learning for pedestrian detection with covariance matrix feature
作者 黃靈威
Huang, Ling Wei
貢獻者 廖文宏
Liao, Wen Hung
黃靈威
Huang, Ling Wei
關鍵詞 半監督式學習
支持向量機
單純貝氏分類器
共變異描述子
Semi-supervised learning
Support vector machine
Naïve Bayes classifier
Covariance descriptor
日期 2008
上傳時間 8-Dec-2010 12:06:40 (UTC+8)
摘要 行人偵測為物件偵測領域中一個極具挑戰性的議題。其主要問題在於人體姿勢以及衣著服飾的多變性,加之以光源照射狀況迥異,大幅增加了辨識的困難度。吾人在本論文中提出利用共變異矩陣描述子及結合單純貝氏分類器與級聯支持向量機的線上學習辨識器,以增進行人辨識之正確率與重現率。
實驗結果顯示,本論文所提出之線上學習策略在某些辨識狀況較差之資料集中能有效提升正確率與重現率達百分之十四。此外,即便於相同之初始訓練條件下,在USC Pedestrian Detection Test Set、 INRIA Person dataset 及 Penn-Fudan Database for Pedestrian Detection and Segmentation三個資料集中,本研究之正確率與重現率亦較HOG搭配AdaBoost之行人辨識方式為優。
Pedestrian detection is an important yet challenging problem in object classification due to flexible body pose, loose clothing and ever-changing illumination. In this thesis, we employ covariance feature and propose an on-line learning classifier which combines naïve Bayes classifier and cascade support vector machine (SVM) to improve the precision and recall rate of pedestrian detection in a still image.

Experimental results show that our on-line learning strategy can improve precision and recall rate about 14% in some difficult situations. Furthermore, even under the same initial training condition, our method outperforms HOG + AdaBoost in USC Pedestrian Detection Test Set, INRIA Person dataset and Penn-Fudan Database for Pedestrian Detection and Segmentation.
參考文獻 [1] IBM Smart Surveillance System http://www.research.ibm.com/peoplevision/
[2] オムロン株式会社公開特許公報【移動体検出方法及び装置並びに移動体認識方法及び装置並びに人間検出方法及び装置】日本国特許庁,1999
[3] Oncel Tuzel, Fatih Porikli, and Peter Meer, “Human detection via classification on Riemannian manifolds”, IEEE Computer Society International Conference on Computer Vision and Pattern Recognition, pp. 1–8, 2007.
[4] Arsigny Vincent, Fillard Pierre, Pennec Xavier, Ayache Nicholas, “Geometric means in a novel vector space structure on symmetric positive-definite matrices”, SIAM Journal on Matrix Analysis and Applications, Vol. 29(1), pp. 328–347, 2006.
[5] Wolfgang Förstner, Boudewijn Moonen, “A metric for covariance matrices”, Technical report, Stuttgart University, Dept. of Geodesy and Geoinformatics, 1999.
[6] Christopher Wren, Ali Azarbayejani, Trevor Darrell, Alex Pentland, “Pfinder: real-time tracking of the human body”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19(7), pp. 780–785, 1997.
[7] Csaba Beleznai , Bernhard Fruhstuck, Horst Bischof, “Human detection in groups using a fast mean shift procedure”, International Conference on Image Processing, Vol. 1, pp. 349–352, 2004.
[8] Tetsuji Haga, Kazuhiko Sumi, Yasushi Yagi, “Human detection in outdoor scene using spatio-temporal motion analysis”, Proceedings of the 17th International Conference on Pattern Recognition, Vol. 4, pp. 331–334, 2004.
[9] How-Lung Eng, Junxian Wang, Alvin H. Kam, Wei-Yun Yau, “A Bayesian framework for robust human detection and occlusion handling using human shape model”, Proceedings of the 17th International Conference on Pattern Recognition, pp. 257 – 260, 2004.
[10] Constantine P. Papageorgiou, Michael Oren, Tomaso Poggio, “A general framework for object detection”, Proceedings of the 6th International Conference on Computer Vision, pp. 555–562,1998.
[11] Paul Viola, Michael J. Jones, Daniel Snow, “Detecting pedestrians using patterns of motion and appearance”, Proceedings of the 9th International Conference on Computer Vision, Vol. 2, pp. 734–741, 2003.
[12] Navneet Dalal, Bill Triggs, “Histograms of oriented gradients for human detection”, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 886–893,2005.
[13] Fatih Porikli, Oncel Tuzel, Peter Meer, “Covariance tracking using model update based on lie algebra” , Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,, pp. 728–735, 2006.
[14] Shumeet Baluja, “Probabilistic modeling for face orientation discrimination: learning from labeled and unlabeled data”, Neural Information Processing Systems, pp. 854–860, 1998.
[15] Kanal Paul Nigam, “Using unlabeled data to improve text classification (Technical Report CMU-CS-01-126)”, Carnegie Mellon University. Doctoral Dissertation, pp.27 2001.
[16] Alex D. Holub, Pietro Perona, Max Welling, “Exploiting unlabeled data for hybrid object classification”, NIPS Workshop in Inter-Class Transfer, 2005.
[17] Yuanqing Li, Cuntai Guan, Huiqi Li, Zhengyang Chin, “A self-training semi-supervised SVM algorithm and its application in an EEG-based brain computer interface speller system”, Pattern Recognition Letters, Vol. 29(9), pp. 1285-1294, 2008.
[18] Raghav Subbarao, Peter Meer, “Nonlinear mean shift for clustering over analytic manifolds”, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1168–1175, 2006.
[19] Fatih Porikli, Tekin Kocak, "Robust license plate detection using covariance descriptor in a neural network framework," IEEE International Conference on Advanced Video and Signal Based Surveillance, pp.107, 2006
[20] Oncel Tuzel, Fatih Porikli, Peter Meer, “Region covariance: A fast descriptor for detection and classification”, Proceedings of the 9th European Conference on Computer Vision, Vol. 2, pp. 589–600, 2006.
[21] Robert Gilmore, “Lie groups, Lie algebras, and some of their applications” , pp. 77 ,Dover, 2002
[22] Xavier Pennec, Pierre Fillard, Nicholas Ayache, ”A Riemannian framework for tensor computing”, International Journal of Computer Vision, pp.41-66, 2006.
[23] Jonathan H. Manton, “A centroid (Karcher mean) approach to the joint approximate diagonalisation problem: The real symmetric case”, Digital Signal Processing, Vol.16, pp. 468-478, 2006.
[24] Chui-Yu Chiu, Yuan-Ting Huang. “Integration of support vector machine with naïve Bayesian classifier for spam classification”, Fuzzy Systems and Knowledge Discovery 4th International Conference, Vol. 1, pp. 24-27, 2007.
[25] http://cbcl.mit.edu/software-datasets/PedestrianData.html
[26] http://www.science.uva.nl/research/isla/downloads/pedestrians/index.html
[27] http://pascal.inrialpes.fr/data/human/
[28] http://iris.usc.edu/~bowu/DatasetWebpage/dataset.html
[29] http://www.cis.upenn.edu/~jshi/ped_html/.
[30] Ivan Laptev, "Improvements of object detection using boosted histograms", Proceedings of the 17th British Machine Vision Conference, pp. III:949-958, 2006. http://www.irisa.fr/vista/Equipe/People/Laptev/download.html
[31] http://pascallin.ecs.soton.ac.uk/challenges/VOC/voc2006/index.html
[32] Sakrapee Paisitkriangkrai, Chunhua Shen, Jian Zhang, ”An experimental evaluation of local features for pedestrian classification”, Digital Image Computing Techniques and Applications, 9th Biennial Conference of the Australian Pattern Recognition Society, pp.53-60, 2007
描述 碩士
國立政治大學
資訊科學學系
96971006
97
資料來源 http://thesis.lib.nccu.edu.tw/record/#G0096971006
資料類型 thesis
dc.contributor.advisor 廖文宏zh_TW
dc.contributor.advisor Liao, Wen Hungen_US
dc.contributor.author (Authors) 黃靈威zh_TW
dc.contributor.author (Authors) Huang, Ling Weien_US
dc.creator (作者) 黃靈威zh_TW
dc.creator (作者) Huang, Ling Weien_US
dc.date (日期) 2008en_US
dc.date.accessioned 8-Dec-2010 12:06:40 (UTC+8)-
dc.date.available 8-Dec-2010 12:06:40 (UTC+8)-
dc.date.issued (上傳時間) 8-Dec-2010 12:06:40 (UTC+8)-
dc.identifier (Other Identifiers) G0096971006en_US
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/49471-
dc.description (描述) 碩士zh_TW
dc.description (描述) 國立政治大學zh_TW
dc.description (描述) 資訊科學學系zh_TW
dc.description (描述) 96971006zh_TW
dc.description (描述) 97zh_TW
dc.description.abstract (摘要) 行人偵測為物件偵測領域中一個極具挑戰性的議題。其主要問題在於人體姿勢以及衣著服飾的多變性,加之以光源照射狀況迥異,大幅增加了辨識的困難度。吾人在本論文中提出利用共變異矩陣描述子及結合單純貝氏分類器與級聯支持向量機的線上學習辨識器,以增進行人辨識之正確率與重現率。
實驗結果顯示,本論文所提出之線上學習策略在某些辨識狀況較差之資料集中能有效提升正確率與重現率達百分之十四。此外,即便於相同之初始訓練條件下,在USC Pedestrian Detection Test Set、 INRIA Person dataset 及 Penn-Fudan Database for Pedestrian Detection and Segmentation三個資料集中,本研究之正確率與重現率亦較HOG搭配AdaBoost之行人辨識方式為優。
zh_TW
dc.description.abstract (摘要) Pedestrian detection is an important yet challenging problem in object classification due to flexible body pose, loose clothing and ever-changing illumination. In this thesis, we employ covariance feature and propose an on-line learning classifier which combines naïve Bayes classifier and cascade support vector machine (SVM) to improve the precision and recall rate of pedestrian detection in a still image.

Experimental results show that our on-line learning strategy can improve precision and recall rate about 14% in some difficult situations. Furthermore, even under the same initial training condition, our method outperforms HOG + AdaBoost in USC Pedestrian Detection Test Set, INRIA Person dataset and Penn-Fudan Database for Pedestrian Detection and Segmentation.
en_US
dc.description.tableofcontents 第一章 研究目的 1
第二章 簡介與文獻回顧 3
第三章 共變異矩陣描述子 7
3.1 利用積分圖像計算任意矩形之共變異矩陣描述子 8
3.2 利用李群表示描述子間之運算 11
3.3 指數映射與對數映射 12
第四章 在切空間中使用SVM 15
第五章 行人辨識系統 20
5.1 行人辨識系統架構 20
5.2 訓練資料與測試資料來源及訓練方式 22
5.3 偵測視窗之整理合併 31
第六章 實作、訓練與辨識結果 33
6.1 實作 33
6.2 訓練與辨識結果 37
6.3 與其他描述子及辨識器之比較 40
6.4 線上學習後辨識能力增減狀況 43
第七章 結論與後續研究改進方向 56
參考文獻 58
zh_TW
dc.format.extent 96780 bytes-
dc.format.extent 107324 bytes-
dc.format.extent 110318 bytes-
dc.format.extent 156398 bytes-
dc.format.extent 177428 bytes-
dc.format.extent 295817 bytes-
dc.format.extent 206811 bytes-
dc.format.extent 1318138 bytes-
dc.format.extent 920755 bytes-
dc.format.extent 178664 bytes-
dc.format.extent 146913 bytes-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.format.mimetype application/pdf-
dc.language.iso en_US-
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0096971006en_US
dc.subject (關鍵詞) 半監督式學習zh_TW
dc.subject (關鍵詞) 支持向量機zh_TW
dc.subject (關鍵詞) 單純貝氏分類器zh_TW
dc.subject (關鍵詞) 共變異描述子zh_TW
dc.subject (關鍵詞) Semi-supervised learningen_US
dc.subject (關鍵詞) Support vector machineen_US
dc.subject (關鍵詞) Naïve Bayes classifieren_US
dc.subject (關鍵詞) Covariance descriptoren_US
dc.title (題名) 應用共變異矩陣描述子及半監督式學習於行人偵測zh_TW
dc.title (題名) Semi-supervised learning for pedestrian detection with covariance matrix featureen_US
dc.type (資料類型) thesisen
dc.relation.reference (參考文獻) [1] IBM Smart Surveillance System http://www.research.ibm.com/peoplevision/zh_TW
dc.relation.reference (參考文獻) [2] オムロン株式会社公開特許公報【移動体検出方法及び装置並びに移動体認識方法及び装置並びに人間検出方法及び装置】日本国特許庁,1999zh_TW
dc.relation.reference (參考文獻) [3] Oncel Tuzel, Fatih Porikli, and Peter Meer, “Human detection via classification on Riemannian manifolds”, IEEE Computer Society International Conference on Computer Vision and Pattern Recognition, pp. 1–8, 2007.zh_TW
dc.relation.reference (參考文獻) [4] Arsigny Vincent, Fillard Pierre, Pennec Xavier, Ayache Nicholas, “Geometric means in a novel vector space structure on symmetric positive-definite matrices”, SIAM Journal on Matrix Analysis and Applications, Vol. 29(1), pp. 328–347, 2006.zh_TW
dc.relation.reference (參考文獻) [5] Wolfgang Förstner, Boudewijn Moonen, “A metric for covariance matrices”, Technical report, Stuttgart University, Dept. of Geodesy and Geoinformatics, 1999.zh_TW
dc.relation.reference (參考文獻) [6] Christopher Wren, Ali Azarbayejani, Trevor Darrell, Alex Pentland, “Pfinder: real-time tracking of the human body”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19(7), pp. 780–785, 1997.zh_TW
dc.relation.reference (參考文獻) [7] Csaba Beleznai , Bernhard Fruhstuck, Horst Bischof, “Human detection in groups using a fast mean shift procedure”, International Conference on Image Processing, Vol. 1, pp. 349–352, 2004.zh_TW
dc.relation.reference (參考文獻) [8] Tetsuji Haga, Kazuhiko Sumi, Yasushi Yagi, “Human detection in outdoor scene using spatio-temporal motion analysis”, Proceedings of the 17th International Conference on Pattern Recognition, Vol. 4, pp. 331–334, 2004.zh_TW
dc.relation.reference (參考文獻) [9] How-Lung Eng, Junxian Wang, Alvin H. Kam, Wei-Yun Yau, “A Bayesian framework for robust human detection and occlusion handling using human shape model”, Proceedings of the 17th International Conference on Pattern Recognition, pp. 257 – 260, 2004.zh_TW
dc.relation.reference (參考文獻) [10] Constantine P. Papageorgiou, Michael Oren, Tomaso Poggio, “A general framework for object detection”, Proceedings of the 6th International Conference on Computer Vision, pp. 555–562,1998.zh_TW
dc.relation.reference (參考文獻) [11] Paul Viola, Michael J. Jones, Daniel Snow, “Detecting pedestrians using patterns of motion and appearance”, Proceedings of the 9th International Conference on Computer Vision, Vol. 2, pp. 734–741, 2003.zh_TW
dc.relation.reference (參考文獻) [12] Navneet Dalal, Bill Triggs, “Histograms of oriented gradients for human detection”, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 886–893,2005.zh_TW
dc.relation.reference (參考文獻) [13] Fatih Porikli, Oncel Tuzel, Peter Meer, “Covariance tracking using model update based on lie algebra” , Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,, pp. 728–735, 2006.zh_TW
dc.relation.reference (參考文獻) [14] Shumeet Baluja, “Probabilistic modeling for face orientation discrimination: learning from labeled and unlabeled data”, Neural Information Processing Systems, pp. 854–860, 1998.zh_TW
dc.relation.reference (參考文獻) [15] Kanal Paul Nigam, “Using unlabeled data to improve text classification (Technical Report CMU-CS-01-126)”, Carnegie Mellon University. Doctoral Dissertation, pp.27 2001.zh_TW
dc.relation.reference (參考文獻) [16] Alex D. Holub, Pietro Perona, Max Welling, “Exploiting unlabeled data for hybrid object classification”, NIPS Workshop in Inter-Class Transfer, 2005.zh_TW
dc.relation.reference (參考文獻) [17] Yuanqing Li, Cuntai Guan, Huiqi Li, Zhengyang Chin, “A self-training semi-supervised SVM algorithm and its application in an EEG-based brain computer interface speller system”, Pattern Recognition Letters, Vol. 29(9), pp. 1285-1294, 2008.zh_TW
dc.relation.reference (參考文獻) [18] Raghav Subbarao, Peter Meer, “Nonlinear mean shift for clustering over analytic manifolds”, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1168–1175, 2006.zh_TW
dc.relation.reference (參考文獻) [19] Fatih Porikli, Tekin Kocak, "Robust license plate detection using covariance descriptor in a neural network framework," IEEE International Conference on Advanced Video and Signal Based Surveillance, pp.107, 2006zh_TW
dc.relation.reference (參考文獻) [20] Oncel Tuzel, Fatih Porikli, Peter Meer, “Region covariance: A fast descriptor for detection and classification”, Proceedings of the 9th European Conference on Computer Vision, Vol. 2, pp. 589–600, 2006.zh_TW
dc.relation.reference (參考文獻) [21] Robert Gilmore, “Lie groups, Lie algebras, and some of their applications” , pp. 77 ,Dover, 2002zh_TW
dc.relation.reference (參考文獻) [22] Xavier Pennec, Pierre Fillard, Nicholas Ayache, ”A Riemannian framework for tensor computing”, International Journal of Computer Vision, pp.41-66, 2006.zh_TW
dc.relation.reference (參考文獻) [23] Jonathan H. Manton, “A centroid (Karcher mean) approach to the joint approximate diagonalisation problem: The real symmetric case”, Digital Signal Processing, Vol.16, pp. 468-478, 2006.zh_TW
dc.relation.reference (參考文獻) [24] Chui-Yu Chiu, Yuan-Ting Huang. “Integration of support vector machine with naïve Bayesian classifier for spam classification”, Fuzzy Systems and Knowledge Discovery 4th International Conference, Vol. 1, pp. 24-27, 2007.zh_TW
dc.relation.reference (參考文獻) [25] http://cbcl.mit.edu/software-datasets/PedestrianData.htmlzh_TW
dc.relation.reference (參考文獻) [26] http://www.science.uva.nl/research/isla/downloads/pedestrians/index.htmlzh_TW
dc.relation.reference (參考文獻) [27] http://pascal.inrialpes.fr/data/human/zh_TW
dc.relation.reference (參考文獻) [28] http://iris.usc.edu/~bowu/DatasetWebpage/dataset.htmlzh_TW
dc.relation.reference (參考文獻) [29] http://www.cis.upenn.edu/~jshi/ped_html/.zh_TW
dc.relation.reference (參考文獻) [30] Ivan Laptev, "Improvements of object detection using boosted histograms", Proceedings of the 17th British Machine Vision Conference, pp. III:949-958, 2006. http://www.irisa.fr/vista/Equipe/People/Laptev/download.htmlzh_TW
dc.relation.reference (參考文獻) [31] http://pascallin.ecs.soton.ac.uk/challenges/VOC/voc2006/index.htmlzh_TW
dc.relation.reference (參考文獻) [32] Sakrapee Paisitkriangkrai, Chunhua Shen, Jian Zhang, ”An experimental evaluation of local features for pedestrian classification”, Digital Image Computing Techniques and Applications, 9th Biennial Conference of the Australian Pattern Recognition Society, pp.53-60, 2007zh_TW