Structured Output Prediction with Hierarchical Loss Functions for Seafloor Imagery Taxonomic Categorization

In this paper we study the challenging problem of seafloor imagery taxonomic categorization. Our contribution is threefold. First, we demonstrate that this task can be elegantly translated into a Structured SVM learning framework. Second, we introduce a taxonomic loss function in the structured output classification objective during learning that is shown to improve the performance over other loss functions. And third, we show how the Structured SVM can naturally deal with the problem of learning from data imbalance by scaling the cost of misclassification during the optimization. We present a thorough experimental evaluation using the challenging and publicly available Tasmania Coral Point Count dataset, where our models drastically outperform the state-of-the-art-results reported.

[1]  Nathalie Japkowicz,et al.  The class imbalance problem: A systematic study , 2002, Intell. Data Anal..

[2]  Stefan B. Williams,et al.  Monitoring of Benthic Reference Sites: Using an Autonomous Underwater Vehicle , 2012, IEEE Robotics & Automation Magazine.

[3]  JapkowiczNathalie,et al.  The class imbalance problem: A systematic study , 2002 .

[4]  S. Kumar,et al.  Segmentation and Classification of Coral for Oceanographic Surveys: A Semi-Supervised Machine Learning Approach , 2006, OCEANS 2006 - Asia Pacific.

[5]  Nello Cristianini,et al.  Controlling the Sensitivity of Support Vector Machines , 1999 .

[6]  Thomas Hofmann,et al.  Hierarchical document categorization with support vector machines , 2004, CIKM '04.

[7]  Thomas Hofmann,et al.  Support vector machine learning for interdependent and structured output spaces , 2004, ICML.

[8]  O. Pizarro,et al.  Automated species detection: An experimental approach to kelp detection from sea-oor AUV images , 2012 .

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

[10]  C. Saloma,et al.  Image classification of coral reef components from underwater color video , 2001, MTS/IEEE Oceans 2001. An Ocean Odyssey. Conference Proceedings (IEEE Cat. No.01CH37295).

[11]  David J. Kriegman,et al.  Automated annotation of coral reef survey images , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[12]  Ke Wang,et al.  Building Hierarchical Classifiers Using Class Proximity , 1999, VLDB.

[13]  Stefan B. Williams,et al.  Hierarchical Classification in AUV Imagery , 2013, FSR.

[14]  Václav Hlavác,et al.  Multi-class support vector machine , 2002, Object recognition supported by user interaction for service robots.

[15]  Gustavo Camps-Valls,et al.  Structured Output SVM for Remote Sensing Image Classification , 2009, 2009 IEEE International Workshop on Machine Learning for Signal Processing.

[16]  Matti Pietikäinen,et al.  Rotation Invariant Image Description with Local Binary Pattern Histogram Fourier Features , 2009, SCIA.

[17]  John J. Leonard,et al.  Cooperative AUV Navigation Using a Single Surface Craft , 2009, FSR.

[18]  Motoaki Kawanabe,et al.  On Taxonomies for Multi-class Image Categorization , 2012, International Journal of Computer Vision.

[19]  Matthew Dunbabin,et al.  Automated Counting of the Northern Pacific Sea Star in the Derwent Using Shape Recognition , 2007, 9th Biennial Conference of the Australian Pattern Recognition Society on Digital Image Computing Techniques and Applications (DICTA 2007).