logo

SCIENTIA SINICA Informationis, Volume 48, Issue 12: 1670-1680(2018) https://doi.org/10.1360/N112018-00143

Multi-instance multi-label new label learning

More info
  • ReceivedMay 31, 2018
  • AcceptedAug 14, 2018
  • PublishedDec 13, 2018

Abstract

In the studies of traditional multi-label learning, they assume the observed label distribution in the training data is the same as the true label distribution in the test data. In real applications, however, it is not the case; some existing semantics are not labeled throughout the dataset. In the testing phase, the learning system not only requires a good performance on the known target labels but also needs to detect the potential new labels. This paper proposes an end-to-end multi-view multi-instance multi-label learning approach, EM3NL, combining deep learning with multi-view multi-instance multi-label learning framework to address the new label learning problem. The performance on MS-COCO datasets validates the effectiveness of EM3NL on both the performance on the known labels and the new labels.


Funded by

国家自然科学基金(61673201)

国家自然科学基金(61333014)


References

[1] Boutell M R, Luo J, Shen X. Learning multi-label scene classification. Pattern Recognition, 2004, 37: 1757-1771 CrossRef Google Scholar

[2] Zhang M L, Zhang K. Multi-label learning by exploiting label dependency. In: Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Washington, 2010. 999--1008. Google Scholar

[3] Read J, Pfahringer B, Holmes G. Classifier chains for multi-label classification. Mach Learn, 2011, 85: 333-359 CrossRef Google Scholar

[4] Huang S J, Zhou Z H. Multi-label learning by exploiting label correlations locally. In: Proceedings of the 26th AAAI Conference on Artificial Intelligence, Toronto, 2012. 949--955. Google Scholar

[5] Zhang M L, Zhou Z H. A Review on Multi-Label Learning Algorithms. IEEE Trans Knowl Data Eng, 2014, 26: 1819-1837 CrossRef Google Scholar

[6] Zhou Z H, Zhang M L. Multi-Label Learning. Berlin: Springer, 2017. Google Scholar

[7] Zhou Z H. A brief introduction to weakly supervised learning. Natl Sci Rev, 2018, 5: 44-53 CrossRef Google Scholar

[8] Xu M, Jin R, Zhou Z H. Speedup matrix completion with side information: application to multi-label learning. In: Proceedings of Advances in Neural Information Processing Systems, Lake Tahoe, 2013. 2301--2309. Google Scholar

[9] Xu L L, Wang Z, Shen Z F, et al. Learning low-rank label correlations for multi-label classification with missing labels. In: Proceedings of the 14th IEEE International Conference on Data Mining, Shenzhen, 2014. 1067--1072. Google Scholar

[10] Yu H F, Jain P, Kar P, et al. Large-scale multi-label learning with missing labels. In: Proceedings of the 31th International Conference on Machine Learning, Beijing, 2014. 593--601. Google Scholar

[11] Zhu Y, Kwok J T, Zhou Z H. Multi-Label Learning with Global and Local Label Correlation. IEEE Trans Knowl Data Eng, 2018, 30: 1081-1094 CrossRef Google Scholar

[12] Zhu Y, Ting K M, Zhou Z H. Discover multiple novel labels in multi-instance multi-label learning. In: Proceedings of the 31th AAAI Conference on Artificial Intelligence, San Francisco, 2017. 2977--2984. Google Scholar

[13] Zhu Y, Ting K M, Zhou Z H. Multi-label learning with emerging new labels. In: Proceedings of the 16th International Conference on Data Mining, Barcelona, 2016. Google Scholar

[14] Zhou Z H, Zhang M L. Multi-instance multi-label learning with application to scene classification. In: Proceedings of Advances in Neural Information Processing Systems, Vancouver, 2007. 1609--1616. Google Scholar

[15] Zhou Z H, Zhang M L, Huang S J. Multi-instance multi-label learning. Artificial Intelligence, 2012, 176: 2291-2320 CrossRef Google Scholar

[16] Zhang M L. A k-nearest neighbor based multi-instance multi-label learning algorithm. In: Proceedings of the 22nd IEEE International Conference on Tools with Artificial Intelligence, Arras, 2010. 207--212. Google Scholar

[17] Briggs F, Fern X Z, Raich R. Rank-loss support instance machines for miml instance annotation. In: Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Beijing, 2012. 534--542. Google Scholar

[18] Huang S J, Gao W, Zhou Z H. Fast multi-instance multi-label learning. In: Proceedings of the 28th AAAI Conference on Artificial Intelligence, Quebec City, 2014. 1868--1874. Google Scholar

[19] Pham A T, Raich R, Fern X Z. Dynamic Programming for Instance Annotation in Multi-Instance Multi-Label Learning.. IEEE Trans Pattern Anal Mach Intell, 2017, 39: 2381-2394 CrossRef PubMed Google Scholar

[20] Bengio Y, Courville A, Vincent P. Representation learning: a review and new perspectives.. IEEE Trans Pattern Anal Mach Intell, 2013, 35: 1798-1828 CrossRef PubMed Google Scholar

[21] Schmidhuber J. Deep learning in neural networks: an overview.. Neural Networks, 2015, 61: 85-117 CrossRef PubMed Google Scholar

[22] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521: 436-444 CrossRef PubMed ADS Google Scholar

[23] Blum A, Mitchell T. Combining labeled and unlabeled data with co-training. In: Proceedings of the 11th Conference on Computational Learning Theory, Madison, 1998. 92--100. Google Scholar

[24] Wang W, Zhou Z H. A new analysis of co-training. In: Proceedings of the 27th International Conference on Machine Learning, Haifa, 2010. 1135--1142. Google Scholar

[25] Guo Y H, Xiao M. Cross language text classification via subspace co-regularized multi-view learning. In: Proceedings of the 29th International Conference on Machine Learning, Edinburgh, 2012. 1615--1622. Google Scholar

[26] Guo Y H. Convex subspace representation learning from multi-view data. In: Proceedings of the 27th AAAI Conference on Artificial Intelligence, Bellevue, 2013. 387--393. Google Scholar

[27] Zhu Y, Gao W, Zhou Z H. One-pass multi-view learning. In: Proceedings of the 7th Asian Conference on Machine Learning, Hong Kong, 2015. 407--422. Google Scholar

[28] Feng J, Zhou Z H. Deep MIML network. In: Proceedings of the 31th AAAI Conference on Artificial Intelligence, San Francisco, 2017. 1884--1890. Google Scholar

[29] Pham A T, Raich R, Fern X Z, et al. Multi-instance multi-label learning in the presence of novel class instances. In: Proceedings of the 32nd International Conference on Machine Learning, Lille, 2015. 2427--2435. Google Scholar

[30] Lin T Y, Maire M, Belongie S J, et al. Microsoft COCO: common objects in context. In: Proceedings of the 13th European Conference on Computer Vision, 2014. 740--755. Google Scholar

[31] Zeiler M D. Adadelta: an adaptive learning rate method. 2012,. arXiv Google Scholar

Copyright 2019 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有

京ICP备18024590号-1