显著性轮廓提取、轮廓编组综述要看的内容（一）Berkeley学派

Berkeley 学派

Jitendra Malik

不用介绍了吧，太大的牛了。如果要介绍起来，可要花不少篇幅了。研究兴趣：分割，知觉组织，纹理，立体视觉，识别等。他带的 26 个学生都在学术和工业领域有重要影响力。

" Contour Detection and Hierarchical Image Segmentation "

P. Arbelaez, M. Maire, C. Fowlkes, J. Malik

IEEE Trans. on PAMI 2010 (in press)

paper

 

"Learning Probabilistic Models for Contour Completion in Natural Images "

X. Ren, C.C. Fowlkes and J. Malik

International Journal of Computer Vision , 77:47-63, 2008

paper

 

" Local Figure/Ground Cues are Valid for Natural Images, "

C.C. Fowlkes, D.R. Martin, J. Malik

Journal of Vision, 7(8):2, 1-9, http://journalofvision.org/7/8/2/, 2007

paper

 

" Learning to Detect Natural Image Boundaries Using Local Brightness, Color and Texture Cues"

David Martin, Charless Fowlkes and Jitendra Malik

IEEE Trans. on Pattern Analysis and Machine Intelligence , 26(5), 530-549, May 2004.

paper

 

" Contour and Texture Analysis for Image Segmentation"

Jitendra Malik, Serge Belongie, Thomas Leung and Jianbo Shi

International Journal of Computer Vision , 43(1), 7-27, June 2001.

paper

"Normalized Cuts and Image Segmentation"

Jianbo Shi and Jitendra Malik

IEEE Transactions on Pattern Analysis and Machine Intelligence , 22(8), 888-905, August 2000.

Paper

"Image Recognition: Visual Grouping, Recognition and Learning"

Joachim M. Buhmann, Jitendra Malik and Pietro Perona

Proc. of National Academy of Sciences , 96(25), 14203-14204, Dec 1999.

Paper

Shi Jianbo

又一个华人大牛！是 Malik 98 年的博士，论文题目是： Perceptual Organization and Image Segmentation ，做的是 Normalized Cuts 。 N-Cuts 也奠定了他在 视觉学术领域谱聚类算法研究方面的地位。在卡耐基梅隆做过研究，现在是宾夕法尼亚大学 A.P. 。

研究兴趣是：识别、分割、行为分析。现在做的偏向高层了。

他相信， grouping 是感知的通用计算引擎，如何指导 grouping 过程来达到特定的视觉任务是他们关心的问题。

重点看 http://www.cis.upenn.edu/~jshi/grouping-thoughts.html 。虽然没有特别考虑 contour ，但看 Shi 的团队在 grouping 问题上根本思想是什么，并且将其思想用在 contour 上是很有意义的，也是可能创新的地方所在。

Image Segmentation and Normalized Cuts

Top-down constraints:

Segmentation Given Partial Grouping Constraints

Stella X. Yu and Jianbo Shi

IEEE Transactions on Pattern Analysis and Machine Intelligence, 26(2):173-83, 2004

A preliminary version appeared as:

Grouping with Bias

Stella X. Yu and Jianbo Shi

Neural Information Processing Systems (NIPS'01), Vancouver, BC, December 3-5, 2001

 

Multi-class spectral graph partitioning:

Multiclass Spectral Clustering

Stella X. Yu and Jianbo Shi

International Conference on Computer Vision, Nice, France, 11-17 Oct 2003

 

Random walk view of segmentation, and learning spectral graph partitioning:

Learning Segmentation with Random Walk

Marina Maila and Jianbo Shi

Neural Information Processing Systems, NIPS, 2001

 

A Random Walks View of Spectral Segmentation

Marina Maila and Jianbo Shi

AI and STATISTICS 2001(AISTATS ) 2001

 

Graph with negative weights, and asymmetric weights:

Understanding Popout through Repulsion

Stella X. Yu and Jianbo Shi

IEEE Conf. On Computer Vision and Pattern Recognition (CVPR), Kauai, Hawaii, December 11-13, 2001

 

Segmentation with Directed Relationships

Stella X. Yu and Jianbo Shi

Third International Workshop on Energy Minimization Methods in Computer Vision and Pattern Recognition(EMMCVPR'01 ), INRIASophia-Antipolis, France, September 3 - 5, 2001

 

Segmentation with Pairwise Attraction and Repulsion

Stella X. Yu and Jianbo Shi

International Conference on Computer Vision(ICCV), 2001

 

Nomalized Cuts for Image segmentation:

Contour and Texture Analysis for Image Segmentation

Jitendra Malik Serge Belongie Thomas Leung and Jianbo Shi

International Journal of Computer Vision 2001

A preliminary version appeared as

Textons, Contours and Regions: Cue Integration in Image Segmentation

Jitendra Malik Serge Belongie Thomas Leung and Jianbo Shi

Int. Conf. Computer Vision, Sept 1999(ICCV ), Corfu, Greece, 1999

 

Normalized Cuts and Image Segmentation

Jianbo Shi and Jitendra Malik

IEEE Transactions on Pattern Analysis and Machine Intelligence(PAMI ) 2000

A preliminary version appeared as

Normalized Cuts and Image Segmentation

Jianbo Shi and Jitendra Malik

IEEE Conf. Computer Vision and Pattern Recognition(CVPR ), June 1997, Puerto Rico

 

Self Inducing Relational Distance and its Application to Image Segmentation

Jianbo Shi and Jitendra Malik

In Fifth European Conference on Computer Vision(ECCV ), June 1998

Charless C. Fowlkes

原来在 UC Berkeley ，现在 UC Irvine 。 05 年博士毕业，论文题目是： Measuring the Ecological Validity of Grouping and Figure-Ground Cues 。 做的是 buttom-up 的 segmentation 跟 top-down 的 recognition 相结合方面的研究。 brightness 、 color 、 texture ，这些 visual cues 都有研究。这也是 UC 学派的典型研究特点。

 

P. Arbelaez, M. Maire, C. Fowlkes, J. Malik, "Contour Detection and Hierarchical Image Segmentation", TPAMI , in press (2010) [pdf] [web page]

J. Burge, C. Fowlkes, M. Banks, ``Natural-scene statistics predict how the figure-ground cue of convexity affects human depth perception'', Journal of Neuroscience , 30(21):7269-7280 [pdf] [supplement]

X. Ren, C. Fowlkes, J. Malik. "Learning Probabilistic Models for Contour Completion in Natural Images", IJCV (2008), 77:47-63. [pdf]

C. Fowlkes, D. Martin, J. Malik. "Local Figure/Ground Cues are Valid for Natural Images" Journal of Vision , 7(8):2, 1-9. [pdf] [dataset]

X. Ren, C. Fowlkes, J. Malik. "Figure/Ground Assignment in Natural Images", ECCV , Graz, Austria, (May 2006). [pdf]

X. Ren, C. Fowlkes, J. Malik. "Cue Integration for Figure/Ground Labeling", NIPS , Vancouver, Canada, (Dec. 2005). [pdf]

X. Ren, C. Fowlkes, J. Malik. "Scale-Invariant Contour Completion using Conditional Random Fields", ICCV , Beijing, China, (Oct. 2005). [pdf]

X. Ren, C. Fowlkes, J. Malik. "Mid-level Cues Improve Boundary Detection", Technical Report CSD-05-1382, Division of Computer Science, University of California, Berkeley, (March 2005). [pdf]

C. Fowlkes "A Note on Planar Factor Graphs", (Oct. 2004). [pdf]

C. Fowlkes, J. Malik. "How Much Does Globalization Help Segmentation?", Technical Report CSD-04-1340, Division of Computer Science, University of California, Berkeley, (July 2004). [pdf]

D. Martin, C. Fowlkes, J. Malik. "Learning to Detect Natural Image Boundaries Using Local Brightness, Color and Texture Cues", TPAMI 26 (5) p.530-549 [pdf]

C. Fowlkes, D. Martin, J. Malik. "Learning Affinity Functions for Image Segmentation: Combining Patch-based and Gradient-based Approaches", CVPR , Madison, WI, (June 2003). [pdf]

D. Martin, C. Fowlkes, J. Malik. "Learning to Detect Natural Image Boundaries Using Brightness and Texture", NIPS , Vancouver, (Dec 2002). [pdf]

Ren Xiaofeng

本科是在浙大读的， 06 年从 berkeley 博士毕业，师从大牛 Jitendra Malik ，所以，是 Fowlkes 的师弟。博士论文是： Probabilistic Models for Mid-Level Vision 。现在 是华盛顿大学的 affiliate A.P.( 不知道跟 A.P. 有何不同 ) ，西雅图 Intel 实验室的研究科学家。

研究兴趣包括图像特征、编组、分割，目标识别，食品分析。现在做的是 RGB-D 数据（ kinect hacking ）的利用。做过视频中行人的检测跟踪，目标轮廓提取， superpixel 。

· Multi-Scale Improves Boundary Detection in Natural Images.   [abstract] [pdf]

  Xiaofeng Ren, in ECCV '08 , Marseille, 2008.

·   Local Grouping for Optical Flow.   [abstract] [pdf]

  Xiaofeng Ren, in CVPR '08 , Anchorage, 2008.

·   Tracking as Repeated Figure/Ground Segmentation.   [abstract] [pdf] [ps] [bibtex] [video]

  Xiaofeng Ren and Jitendra Malik , in CVPR '07 , Minneapolis 2007.

·   Figure/Ground Assignment in Natural Images.   [abstract] [pdf] [ps] [bibtex]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , in ECCV '06 , volume 2, pages 614-627, Graz 2006.

·   Learning Probabilistic Models for Contour Completion in Natural Images.   [pdf]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , in IJCV Special Issue on Machine Learning for Vision , May 2008.

·   Cue Integration in Figure/Ground Labeling.   [abstract] [pdf] [bibtex]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , in NIPS '05 , Vancouver 2005.

·   Scale-Invariant Contour Completion using Conditional Random Fields.   [abstract] [pdf] [ps] [bibtex]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , in ICCV '05 , volume 2, pages 1214-1221, Beijing 2005.

·   Familiar Configuration Enables Figure/Ground Assignment in Natural Scenes.   [abstract] [poster] [bibtex]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , in VSS 05 , Sarasota, FL 2005.

·   Mid-level Cues Improve Boundary Detection.   [abstract] [pdf] [ps] [bibtex]

  Xiaofeng Ren, Charless Fowlkes and Jitendra Malik , Berkeley Technical Report 05-1382, CSD 2005.

·   The Ecological Statistics of Good Continuation: Multi-scale Markov Models for Contours.   [abstract] [talk][bibtex]

  Xiaofeng Ren and Jitendra Malik , in VSS 02 , Sarasota, FL 2002.

·   A Probabilistic Multi-scale Model for Contour Completion Based on Image Statistics.   [abstract] [pdf] [ps][bibtex]

  Xiaofeng Ren and Jitendra Malik , in ECCV '02 , volume 1, pages 312-327, Copenhagen 2002.

Michael Maire

09 年从 berkeley 毕业，师从 Jitendra Malik ，博士论文题目： Contour Detection and Image Segmentation。现在在加州理工学院 Computer Vision Lab 做博士后。

【 gPb 分割算法】 Using Contours to Detect and Localize Junctions in Natural Images

Michael Maire, Pablo Arbelaez, Charless Fowlkes, and Jitendra Malik,Computer Vision and Pattern Recognition (CVPR), 2008. [PDF] [Slides] [Software] [Code]

Stella Yu

在卡耐基梅隆读的博士。 Shi Jianbo 03 的博士学生，博士题目是： Computational Models of Perceptual Organization。从 CMU 毕业后到 Berkeley 做博士后， 05 年到波士顿大学做 A.P. 。从事谱聚类方法用于图像分割方向的研究。

我对她的 Multiscale 边缘检测感兴趣。另外， Grouping with bias ， Grouping with repulsion 这些都是谱聚类的方法，看看吧，看能不能用得上。