請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/9928完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 莊永裕(Yung-Yu Chuang) | |
| dc.contributor.author | Yu-Ting Hsieh | en |
| dc.contributor.author | 謝毓庭 | zh_TW |
| dc.date.accessioned | 2021-05-20T20:49:59Z | - |
| dc.date.available | 2008-07-03 | |
| dc.date.available | 2021-05-20T20:49:59Z | - |
| dc.date.copyright | 2008-07-03 | |
| dc.date.issued | 2008 | |
| dc.date.submitted | 2008-06-18 | |
| dc.identifier.citation | [1] A. Bosch, A. Zisserman, and X. Munoz. Image Classification using Random Forests and Ferns. Computer Vision, 2007. ICCV 2007. IEEE 11th International Conference on, 2007.
[2] A. Bosch, A. Zisserman, and X. Munoz. Representing shape with a spatial pyramid kernel. Proceedings of the 6th ACM international conference on Image and video retrieval, pages 401–408, 2007. [3] C. Chang and C. Lin. LIBSVM: a library for support vector machines. Software available at http://www. csie. ntu. edu. tw/˜cjlin/libsvm, 80:604–611, 2001. [4] D. Crandall and D. Huttenlocher. Composite Models of Objects and Scenes for Category Recognition. Computer Vision and Pattern Recognition, 2007. CVPR’07. IEEE Conference on, 2007. [5] L. Fei-Fei, R. Fergus, and P. Perona. Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories. Computer Vision and Image Understanding, 106(1):59–70, 2007. [6] R. Fergus, P. Perona, and A. Zisserman. A sparse object category model for efficient learning and exhaustive recognition. Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on, 1, 2005. [7] S. Fidler and A. Leonardis. Towards Scalable Representations of Object Categories: Learning a Hierarchy of Parts. Computer Vision and Pattern Recognition, 2007. CVPR’07. IEEE Conference on, 2007. [8] K. Grauman and T. Darrell. The Pyramid Match Kernel: Discriminative Classification with Sets of Image Features. Computer Vision, 2005. ICCV 2005. Tenth IEEE International Conference on, 2, 2005. [9] S. Hoi, M. Lyu, and E. Chang. Learning the unified kernel machines for classification. Proceedings of the 12th ACMSIGKDD international conference on Knowledge discovery and data mining, pages 187–196, 2006. [10] V. Kwatra, A. Schodl, I. Essa, G. Turk, and A. Bobick. Graphcut textures: Image and video synthesis using graph cuts. ACM Transactions on Graphics, 22(3):277–286, 2003. [11] S. Lazebnik, C. Schmid, and J. Ponce. Beyond bags of features: Spatial pyramid matching for recognizing natural scene categories. Proc. CVPR, 2(2169-2178):1, 2006. [12] Y. Lin, T. Liu, and C. Fuh. Local Ensemble Kernel Learning for Object Category Recognition. Computer Vision and Pattern Recognition, 2007. CVPR’07. IEEE Conference on, 2007. [13] T. Liu, J. Sun, N. Zheng, X. Tang, and H. Shum. Learning to Detect A Salient Object. Proceedings of IEEE Computer Society Conference on Computer and Vision Pattern Recognition (CVPR), 2007. [14] D. Lowe. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision, 60(2):91–110, 2004. [15] F. Odone, A. Barla, and A. Verri. Building kernels from binary strings for image matching. Image Processing, IEEE Transactions on, 14(2):169–180, 2005. [16] Y.Wu, E. Chang, K. Chang, and J. Smith. Optimal multimodal fusion for multimedia data analysis. Proceedings of the 12th annual ACM international conference on Multimedia, pages 572–579, 2004. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/9928 | - |
| dc.description.abstract | 此篇論文主要的研究，是將影像的背景資訊加入一般物體辨識的流程，以提升其準確率。目前大部分的研究並未將影像的前景物體與背景分開考慮，或者只利用前景的資訊。在這一篇論文中，我們試著加入背景資訊以提過一般物體辨別的準確率。
我們使用一個偵測使用者感興趣區域（Region of Interest）的方法來將影像前景的物體偵測出來。更進一步地，使用者感興趣區域周圍的背景資訊可以用來加強物體識別。由於同一個種類的物體通常會出現在某些特定的場合，我們將由實驗說明加入背景資訊對一般物體辨識率的提升。 另一個很有挑戰性的問題是如果將不同的影像特徵合併使用。我們比較了幾個不同的方法在支持向量機（Support Vector Machine）上的表現。實驗結果顯示這些方法在這個問題上的好壞，與他們能否有效運用背景資訊來加提升辨識率。 | zh_TW |
| dc.description.abstract | This thesis introduces background information to generic object recognition problem to increase the accuracy. Most of works do not divide images to foreground and background part, or only utilize foreground information. In this thesis, we tried to leverage background information to help object recognition.
A region of interest (ROI) detector is used to find the foreground object in images. Focusing on foreground object can reduce noisy features from unrelevant background region. Furthermore, the complement area of ROI can be considered as background context. Since objects in a category usually appear in specific context, we will show that adding background clue can improve the recognition accuracy in our experiment. Another challenge problem is how to use different signals together. We compared several methods of feature fusion for machine learning using SVM. Experiment result shows how well these methods can achieve and whether background information benefit them. | en |
| dc.description.provenance | Made available in DSpace on 2021-05-20T20:49:59Z (GMT). No. of bitstreams: 1 ntu-97-R95922017-1.pdf: 732594 bytes, checksum: c32a20f3bb6914b88a89296dbec69cca (MD5) Previous issue date: 2008 | en |
| dc.description.tableofcontents | Acknowledgments iii
Abstract v List of Figures xi List of Tables xiii Chapter 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Chapter 2 Related Work 5 2.1 Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 ROI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.3 Feature Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Chapter 3 Feature Extraction 9 viii 3.1 Grid of Pyramid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.2 Pyramid of Histogram . . . . . . . . . . . . . . . . . . . . . . . . . . 11 3.3 Pyramid Match Kernel . . . . . . . . . . . . . . . . . . . . . . . . . 11 3.4 Foreground Representation . . . . . . . . . . . . . . . . . . . . . . . 12 Chapter 4 Region of Interest 15 4.1 ROI for Object Detection . . . . . . . . . . . . . . . . . . . . . . . . 15 4.1.1 Low-level Feature-based Exhaustive Search . . . . . . . . . . 16 4.1.2 Learning-based Detection with Visual Cue . . . . . . . . . . 17 4.2 Apply ROI to Classification Problem . . . . . . . . . . . . . . . . . . 20 4.2.1 Background Representation . . . . . . . . . . . . . . . . . . 21 Chapter 5 Supervised Learning of Categories 23 5.1 Support Vector Machine . . . . . . . . . . . . . . . . . . . . . . . . 23 5.2 Feature Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 5.2.1 Averaged Kernel . . . . . . . . . . . . . . . . . . . . . . . . 24 5.2.2 Ensemble Learning . . . . . . . . . . . . . . . . . . . . . . . 24 5.2.3 Adaptive Grid Search of Weighting . . . . . . . . . . . . . . 26 5.2.4 Super Kernel Fusion . . . . . . . . . . . . . . . . . . . . . . 27 Chapter 6 Experiment 29 6.1 Caltech 101 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 6.2 Feature Extraction in ROI . . . . . . . . . . . . . . . . . . . . . . . . 30 6.3 Feature Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 6.3.1 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 6.4 Example of Result Image . . . . . . . . . . . . . . . . . . . . . . . . 32 Chapter 7 Conclusion 33 Bibliography 34 | |
| dc.language.iso | en | |
| dc.title | 考慮背景資訊之一般物體辨識 | zh_TW |
| dc.title | Utilizing Background Information for Generic Object Recognition | en |
| dc.type | Thesis | |
| dc.date.schoolyear | 96-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 林智仁(Chih-Jen Lin),徐宏民(Winston H. Hsu) | |
| dc.subject.keyword | 一般物體辨識,背景, | zh_TW |
| dc.subject.keyword | Generic object recognition,background, | en |
| dc.relation.page | 35 | |
| dc.rights.note | 同意授權(全球公開) | |
| dc.date.accepted | 2008-06-19 | |
| dc.contributor.author-college | 電機資訊學院 | zh_TW |
| dc.contributor.author-dept | 資訊工程學研究所 | zh_TW |
| 顯示於系所單位： | 資訊工程學系 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-97-1.pdf | 715.42 kB | Adobe PDF | 檢視/開啟 |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。