Extended Margin and Soft Balanced Strategies in Active Learning

Nowadays active learning is gaining increasing interest in computer vision community, especially on images. The most commonly used query strategy framework is uncertainty sampling usually in a pool-based sampling scenario. In this paper we propose two query strategies for image classification under the uncertainty sampling framework, both of them being improvements of existing techniques. The first strategy, so called Extended Margin incorporates all possible class labels to calculate the informativeness values of unlabeled instances. The second strategy is the improvement of the recently published BAL method, so called Soft Balanced approach, where we suggest new final informativeness score from an uncertainty measure and a novel penalty metric. We used least margin criterion for the former and the latter was calculated from the categorical penalty scores by using soft assignment. We conducted experiments on 60 different test image sets, each of them was a randomly selected subset of the Caltech101 image collection. The experiments were performed in an extended active learning environment and the results showed that the Extended Margin outperforms the least margin approach and the Soft Balanced method overcomes all other competitor method.

[1]  Andrew Zisserman,et al.  The devil is in the details: an evaluation of recent feature encoding methods , 2011, BMVC.

[2]  William A. Gale,et al.  A sequential algorithm for training text classifiers , 1994, SIGIR '94.

[3]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[4]  Yi Yang,et al.  Multi-Class Active Learning by Uncertainty Sampling with Diversity Maximization , 2015, International Journal of Computer Vision.

[5]  Maya R. Gupta,et al.  Theory and Use of the EM Algorithm , 2011, Found. Trends Signal Process..

[6]  Chuang-Hua Chueh,et al.  Cross-Domain Opinion Word Identification with Query-By-Committee Active Learning , 2014, TAAI.

[7]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[8]  D. Angluin.  Queries and Concept Learning , 1988 .

[9]  Dávid Papp,et al.  Balanced Active Learning Method for Image Classification , 2017, Acta Cybern..

[10]  J. MacQueen.  Some methods for classification and analysis of multivariate observations , 1967 .

[11]  David A. Cohn,et al.  Training Connectionist Networks with Queries and Selective Sampling , 1989, NIPS.

[12]  Edward Y. Chang,et al.  Support vector machine active learning for image retrieval , 2001, MULTIMEDIA '01.

[13]  Naila Murray,et al.  Revisiting the Fisher vector for fine-grained classification , 2014, Pattern Recognit. Lett..

[14]  Cordelia Schmid,et al.  Scale & Affine Invariant Interest Point Detectors , 2004, International Journal of Computer Vision.

[15]  Florent Perronnin,et al.  Fisher Kernels on Visual Vocabularies for Image Categorization , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[16]  Kristen Grauman,et al.  Large-Scale Live Active Learning: Training Object Detectors with Crawled Data and Crowds , 2011, CVPR 2011.

[17]  Cordelia Schmid,et al.  Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[18]  Thomas Mensink,et al.  Improving the Fisher Kernel for Large-Scale Image Classification , 2010, ECCV.

[19]  Chih-Jen Lin,et al.  Generalized Bradley-Terry Models and Multi-Class Probability Estimates , 2006, J. Mach. Learn. Res..

[20]  Jan Kautz,et al.  Hierarchical Subquery Evaluation for Active Learning on a Graph , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[21]  Bernhard E. Boser,et al.  A training algorithm for optimal margin classifiers , 1992, COLT '92.

[22]  Florent Perronnin,et al.  Large-scale image retrieval with compressed Fisher vectors , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[23]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[24]  Manali Sharma,et al.  Evidence-based uncertainty sampling for active learning , 2016, Data Mining and Knowledge Discovery.

[25]  Pietro Perona,et al.  Learning Generative Visual Models from Few Training Examples: An Incremental Bayesian Approach Tested on 101 Object Categories , 2004, 2004 Conference on Computer Vision and Pattern Recognition Workshop.

[26]  Tsuhan Chen,et al.  An active learning framework for content-based information retrieval , 2002, IEEE Trans. Multim..

[27]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[28]  David A. Cohn,et al.  Improving generalization with active learning , 1994, Machine Learning.

[29]  Jun Zhou,et al.  Maximizing Expected Model Change for Active Learning in Regression , 2013, 2013 IEEE 13th International Conference on Data Mining.

[30]  Burr Settles,et al.  Active Learning , 2012, Synthesis Lectures on Artificial Intelligence and Machine Learning.

[31]  Kazufumi Kaneda,et al.  Image sequence recognition with active learning using uncertainty sampling , 2013, The 2013 International Joint Conference on Neural Networks (IJCNN).