Resources‎ > ‎

Multiple Instance Learning

I summarize some important work on multiple instance learning, and hope it will be useful to someones who want to know about multiple instance learning. I am not intend to list all the related papers because there are too many. But if you found the latest work or any important work I missed, please notify me. I'll update the list.

I am still update this page to link the PDF file from other places. However, if you found any link violates your copyright. Please notify me and I will remove it. I am also preparing to release my implementation of some MI-algorithms in below and will put it here when it is ready.

Last update: May 12, 2014.

Survey and Others:

[1] Z.-H. Zhou. Multiple Instance Learning: A Survey. Technical Report, Nanjing University, 2004. [PDF] A comprehensive survey on MIL work before 2004.
[2] B. Babenko. Multiple Instance Learning: Algorithms and Applications. 2009. [PDF]. A good note on MIL.
[3] S. Ray  and M. Craven. Supervised versus multiple instance learning: An empirical comparison. In ICML 2005. [PDFSingle-Instance Learning (SIL) v.s. Multi-Instance earning(MIL).
[4] J. Foulds and E. Frank. A Review of Multi-Instnace Assumptions. Knowledge Engineering Review, 25(1):1-25, 2010. [PDF A methodological review from the perspective of MIL assumptions.
[5] J. Amores. Multiple Instance Classification: Review, Taxonomy and Comparative Study. Artificial Intelligence, 201:81-105, 2013. [PDFA recent review with comparative results on MIL.

Instance-Based Methods:

[2] W. Li, L. Duan, I. W. Tsang and D. Xu. Co-Labeling: A new multi-view learning approach for ambiguous problems, In ICDM, 2012. [PDF] (Co-Labeling, Multi-view MIL)
[2] W. Li, L. Duan, D. Xu and I. W. Tsang. Text-based Web Image Retrieval using Progressive Multiple Instance Learning, In ICCV, 2011. [PDF] (MIL-CPB, PMIL-CPB)
[3] Y. Han, Q. Tao, and J. Wang. Avoiding false positive in multi-instance learning. In NIPS, 2010. [PDF]
[4] Y.-F. Li, J. T. Kwok, I. W. Tsang, and Z.-H. Zhou. A convex method for locating regions of interest with multi-instance learning. In ECML, 2009. [PDF] (KI-SVM)
[5] Peter V. Gehler and Olivier Chapelle. Deterministic Annealing for Multiple-Instance Learning. AISTATS, 2007. [PDF] (DA-MIL)
[6] S. Andrews, I. Tsochantaridis, and T. Hofmann. Support Vector Machines for Multiple Instance Learning. In NIPS, 2003. [PDF] (mi-SVM, MI-SVM)

Bag-Based Methods:

[1] B. Babenko, N. Verma, P. Dollar and S. Belongie. Multiple Instance Learning with Manifold Bags. In ICML, 2012. [PDF]
[2] Z. Fu, A. Robles-Kelly and J. Zhou. MILIS: Multiple Instance Learning with Instance Selection. T-PAMI, 33(5): 958-977, 2011. [PDF] (an extension of MILES by learning instance prototypes)
[3] M. Li, J. T. Kwok, and B.-L. Lu. Online multiple instance learning with no regret. In CVPR, 2010 [PDF] (an online version of MILES)
[4] B. Babenko, M. Yang, and S. Belongie. Visual tracking with online multiple instance learning. In CVPR, 2009. [PDF]
[5] Z.-H. Zhou, Y.-Y. Sun, and Y.-F. Li. Multi-instance learning by treating instances as non-i.i.d. samples. In ICML 2009. [PDF] (mi-Graph)
[6] R. Bunescu and R. Mooney. Multiple Insance Learning for Sparse Positive Bags. in ICML, 2007. [PDF] (sMIL = SVM + MI-kernels)
[7] Y. Chen, J. Bi, and J.-Z. Wang. MILES: Multiple Instance Learning via Embedded Instance Selection. T-PAMI, 28(12): 1931-1947, 2006. [PDF] (MILES)
[8] P. Viola, J. C. Platt, and Cha Zhang. Multiple instance boosting for object detection. In NIPS, 2005.  [PDF] (MI-Boosting)
[9] Y. Chen and J. Z. Wang. Image categorization by learning and reasoning with regions. JMLR, 5(27):913–939, 2004. [PDF]

Online Learning:

[1] W. Li, L. Duan, I. W. Tsang and D. Xu. Batch Mode Adaptive Multiple Instance Learning for Computer Vision Tasks. In CVPR 2012. [PDF] (BAMIL: a quasi-online version of PMIL-CPB)
[2] M. Li, J. T. Kwok, and B.-L. Lu. Online multiple instance learning with no regret. In CVPR, 2010 [PDF] (an online version of MILES)
[3] B. Babenko, M. Yang, and S. Belongie. Visual tracking with online multiple instance learning. In CVPR, 2009. [PDF] (online MI-Boosting)


Computer Vision Applications:

Image/Video Categorization/Ranking (Image/Video as an instance):

[1] W. Li, L. Duan, D. Xu and I. W. Tsang. Text-based Web Image Retrieval using Progressive Multiple Instance Learning, In ICCV, 2011. [PDF] (image as instance, progressively constrcut good bags)
[2] S. Vijayanarasimhan and K. Grauman. Keywords to visual categories: Multiple-instance learning for weakly supervised object categorization. In CVPR 2008. [PDF] (image as instance, adopt sMIL)
[3] T. Leung, Y. Song, J. Zhang. Handling Label Noise in Video Classification via Multiple Instance Learning. In ICCV 2011. [PDF] (video as instance, adopt MI-Boosting)

Regional Annotation/Object Detection/Action Detection (Image/Video as a bag)

[4] X. Xue, W. Zhang, J. Zhang, B. Wu, J. Fan and Y. Lu. Correlative Multi-Label Multi-Instance Image Annotation. In ICCV, 2011. [PDF]
[5] Y.-F. Li, J. T. Kwok, I. W. Tsang, and Z.-H. Zhou. A convex method for locating regions of interest with multi-instance learning. In ECML, 2009. [PDF]
[6] P. Viola, J. C. Platt, and Cha Zhang. Multiple instance boosting for object detection. In NIPS, 2005. [PDF]

Tracking:

[7] W. Li, L. Duan, I. W. Tsang and D. Xu. Batch Mode Adaptive Multiple Instance Learning for Computer Vision Tasks. In CVPR 2012. [PDF]
[8] M. Li, J. T. Kwok, and B.-L. Lu. Online multiple instance learning with no regret. In CVPR, 2010. [PDF]
[9] B. Babenko, M. Yang, and S. Belongie. Visual tracking with online multiple instance learning. In CVPR, 2009. [PDF]

Early Work:

[1] T. Dietterich, R. Lathrop, and T. Lozano-Perez. Solving the multiple instance problem with axis-parallel rectangles. AI, 89(1–2):31–71, 1997. [PDF]
[2] O. Maron and T. Lozano-P´erez. A framework for multiple-instance learning. In NIPS, 1998 [PDF]
[3] O. Maron and A. L. Ratan. Multiple-instance learning for natural scene classification. In ICML, 1998. [PDF]
[4] T. Gartner, P .A. Flach, A. Kowalczyk, and A. J. Smola. Multi-instance kernels. In ICML, 2002. [PDF]
[5] Q. Zhang, S. A. Goldman, W. Yu, and J. E. Fritts. Content-based image retrieval using multiple-instance learning. In ICML, 2002. [PDF]
[6] Q. Zhang, S. A. Goldman. EM-DD: An improved multiple-instance learning technique. In NIPS, 2001. [PDF]

Toolkits & Datasets

[1] http://www.miproblems.org/. (A useful website which includes some MI benchmark datasets)
[2]
[3]
Comments