Active Classification: Theory and Application to Underwater Inspection

We discuss the problem in which an autonomous vehicle must classify an object based on multiple views. We focus on the active classification setting, where the vehicle controls which views to select to best perform the classification. The problem is formulated as an extension to Bayesian active learning, and we show connections to recent theoretical guarantees in this area. We formally analyze the benefit of acting adaptively as new information becomes available. The analysis leads to a probabilistic algorithm for determining the best views to observe based on information theoretic costs. We validate our approach in two ways, both related to underwater inspection: 3D polyhedra recognition in synthetic depth maps and ship hull inspection with imaging sonar. These tasks encompass both the planning and recognition aspects of the active classification problem. The results demonstrate that actively planning for informative views can reduce the number of necessary views by up to 80 % when compared to passive methods.

[1]  Bernt Schiele,et al.  Transinformation for active object recognition , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[2]  David Casasent,et al.  Feature Space Trajectory Methods for Active Computer Vision , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Sebastian Thrun,et al.  Probabilistic robotics , 2002, CACM.

[4]  Luc Van Gool,et al.  SURF: Speeded Up Robust Features , 2006, ECCV.

[5]  Franz S. Hover,et al.  Planning Complex Inspection Tasks Using Redundant Roadmaps , 2011, ISRR.

[6]  Alex Zelinsky,et al.  Learning OpenCV---Computer Vision with the OpenCV Library (Bradski, G.R. et al.; 2008)[On the Shelf] , 2009, IEEE Robotics & Automation Magazine.

[7]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[8]  V Myers,et al.  A POMDP for multi-view target classification with an autonomous underwater vehicle , 2010, OCEANS 2010 MTS/IEEE SEATTLE.

[9]  Subhashis Banerjee,et al.  Active recognition through next view planning: a survey , 2004, Pattern Recognit..

[10]  J. Vondrák,et al.  Approximating the Stochastic Knapsack Problem: The Benefit of Adaptivity , 2008 .

[11]  Andreas Krause,et al.  Near-Optimal Bayesian Active Learning with Noisy Observations , 2010, NIPS.

[12]  David P. Williams.  On optimal AUV track-spacing for underwater mine detection , 2010, 2010 IEEE International Conference on Robotics and Automation.

[13]  Hugh F. Durrant-Whyte,et al.  A Bayesian Approach to Optimal Sensor Placement , 1990, Int. J. Robotics Res..

[14]  Tara Javidi,et al.  Active M-ary sequential hypothesis testing , 2010, 2010 IEEE International Symposium on Information Theory.

[15]  Andreas Krause,et al.  Adaptive Submodularity: A New Approach to Active Learning and Stochastic Optimization , 2010, COLT 2010.

[16]  Geoffrey A. Hollinger,et al.  Efficient Multi-robot Search for a Moving Target , 2009, Int. J. Robotics Res..

[17]  Frank P. Ferrie,et al.  Entropy-based gaze planning , 2001, Image Vis. Comput..

[18]  Christopher Hunt,et al.  Notes on the OpenSURF Library , 2009 .

[19]  Andreas Krause,et al.  Near-optimal Nonmyopic Value of Information in Graphical Models , 2005, UAI.

[20]  Dorin Comaniciu,et al.  Conditional feature sensitivity: a unifying view on active recognition and feature selection , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[21]  R. Khan,et al.  Sequential Tests of Statistical Hypotheses. , 1972 .

[22]  Joel W. Burdick,et al.  Dynamic sensor planning with stereo for model identification on a mobile platform , 2010, 2010 IEEE International Conference on Robotics and Automation.

[23]  Stefan B. Williams,et al.  Towards autonomous habitat classification using Gaussian Mixture Models , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[24]  Yiannis Aloimonos,et al.  Active vision , 2004, International Journal of Computer Vision.

[25]  David P. Williams.  Bayesian Data Fusion of Multiview Synthetic Aperture Sonar Imagery for Seabed Classification , 2009, IEEE Transactions on Image Processing.

[26]  DenzlerJoachim,et al.  Information Theoretic Sensor Data Selection for Active Object Recognition and State Estimation , 2002 .

[27]  C. Ian Connolly,et al.  The determination of next best views , 1985, Proceedings. 1985 IEEE International Conference on Robotics and Automation.

[28]  Andreas Krause,et al.  Efficient Informative Sensing using Multiple Robots , 2014, J. Artif. Intell. Res..

[29]  Gerald J. Dobeck,et al.  Fusion of multiple quadratic penalty function support vector machines (QPFSVM) for automated sea mine detection and classification , 2002, SPIE Defense + Commercial Sensing.