Huan Li
ABSTRACT
Cross-domain video concept detection is a challenging task due to the distribution difference between the source domain and target domain. In order to avoid expensive labeling the target-domain data, Active Learning can be used to incrementally learn a target classifier by reusing the one in the source domain. It uses a discriminative query strategy and picks the most ambiguous samples to label, which could fail if the distribution difference is too large. In this paper, to deal with large difference in data distributions, we propose a generative query strategy which is then combined with the existing discriminative one to yield a hybrid method. This method adaptively fits the distribution differences and gives a mixture strategy that performs more robustly compared to both single strategies. Experimental results on TRECVID semantic concept detection task demonstrate superior performance of our hybrid method.
- J. Yang, R. Yan and A. Hauptmann. Cross-Domain video concept detection using adaptive SVMs. In Proc. of 15th Annual ACM Int'l Conf. on Multimedia, 2007. Google Scholar
Digital Library
- B. Settles. Active Learning Literature Survey. Computer Sciences Technical Report 1648, University of Wisconsin-Madison, 2009.Google Scholar
- A. Smeaton and P. Over. Trecvid: Benchmarking the effectiveness of information retrieval tasks on digital video. In Proc. of the Int'l Conf. on Image and Video Retrieval, 2003. Google ScholarDigital Library
- S. J. Pan, I. W. Tsang, J. T. Kwok and Q. Yang. Domain adaptation via transfer component analysis. In Proc. of 21st International Joint Conference on Artificial Intelligence, 2009. Google ScholarDigital Library
- W. Jiang, E. Zavesky, S.-F. Chang, and A. C. Loui. Cross-domain learning methods for high-level visual concept classification. In Proc. of the Int'l Conf. on Image Processing, 2008.Google ScholarCross Ref
- S. J. Pan, and Q. Yang. A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 2009. Google ScholarDigital Library
- D. Lewis and W. Gale. A sequential algorithm for training text classifiers. In Proc. of the ACM SIGIR, 1994. Google ScholarDigital Library
- H.S. Seung, M. Opper, and H. Sompolinsky. Query by committee. In Proc. of the ACM Workshop on Computational Learning Theory, 1992. Google ScholarDigital Library
- B. Settles, M. Craven, and S. Ray. Multiple-instance active learning. In Advances in Neural Information Processing Systems, MIT Press, 2008.Google Scholar
- C.J.C.Burges. A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery, 1998. Google ScholarDigital Library
- V. N. Vapnik. Statistical Learning Theory, 1998.Google Scholar
- S. Tong, and D Koller. Support vector machine active learning with applications to text classification. Journal of Machine Learning Research, 2002. Google ScholarDigital Library
- M. R. Naphade, et al.. A light scale concept ontology for multimedia understanding for TRECVID 2005. IBM Research Technical Report, 2005.Google Scholar
- L. Duan, I.W. Tsang, D. Xu, and S.J. Maybank. Domain transfer SVM for video concept detection. In Proc. of 24th IEEE Conf. on Computer Vision and Pattern Recognition, 2009.Google Scholar
Index Terms
- Hybrid active learning for cross-domain video concept detection
Recommendations
Cross-domain video concept detection: A joint discriminative and generative active learning approach
In this work, we study the problem of cross-domain video concept detection, where the distributions of the source and target domains are different. Active learning can be used to iteratively refine a source domain classifier by querying labels for a few ...
Cost‐effective multi‐instance multilabel active learning
AbstractMulti‐instance multi‐label (MIML) Active Learning (M2AL) aims to improve the learner while reducing the cost as much as possible by querying informative labels of complex bags composed of diverse instances. Existing M2AL solutions suffer high ...
Transfer active learning
CIKM '11: Proceedings of the 20th ACM international conference on Information and knowledge managementActive learning traditionally assumes that labeled and unlabeled samples are subject to the same distributions and the goal of an active learner is to label the most informative unlabeled samples. In reality, situations may exist that we may not have ...
Comments