The Fusion of Optical and Orientation Information in a Markovian Framework for 3D Object Retrieval

In this paper we introduce a new 3D object retrieval model inspired by some well-known mechanisms of the human brain: viewer-centric recognition, Markovian estimations, and fusion of information originating from the visual and vestibular subsystems. We have built a Hidden Markov Model (HMM) framework where 2D object views correspond to states, observations are coded by compact edge and color sensitive descriptors, and orientation sensors are used to secure temporal inference by estimating transition probabilities between states. Our first evaluation results, over a database of 100 3D objects, are very encouraging: the fast and memory efficient new method outperformed previous models.

[1]  Yogendra Kumar Jain,et al.  Efficient view based 3-D object retrieval using Hidden Markov Model , 2013 .

[2]  Memane Swati Bhaskar,et al.  Face Recognition Using Singular Value Decomposition and Hidden Markov Model , 2015 .

[3]  László Czúni,et al.  View centered video-based object recognition for lightweight devices , 2016, 2016 International Conference on Systems, Signals and Image Processing (IWSSIP).

[4]  Yaser Sheikh,et al.  Model generation for video-based object recognition , 2006, MM '06.

[5]  Sheng He,et al.  Viewer-Centered Object Representation in the Human Visual System Revealed by Viewpoint Aftereffects , 2005, Neuron.

[6]  László Czúni,et al.  The use of IMUs for video object retrieval in lightweight devices , 2017, J. Vis. Commun. Image Represent..

[7]  I. Biederman.  Recognition-by-components: a theory of human image understanding. , 1987, Psychological review.

[8]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[9]  Ali Farhadi,et al.  XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks , 2016, ECCV.

[10]  S. Edelman,et al.  Orientation dependence in the recognition of familiar and novel views of three-dimensional objects , 1992, Vision Research.

[11]  W. Penny,et al.  Time Scales of Representation in the Human Brain: Weighing Past Information to Predict Future Events , 2011, Front. Hum. Neurosci..

[12]  Yiannis S. Boutalis,et al.  Accurate Image Retrieval Based on Compact Composite Descriptors and Relevance Feedback Information , 2010, Int. J. Pattern Recognit. Artif. Intell..

[13]  Sameer A. Nene,et al.  Columbia Object Image Library (COIL100) , 1996 .

[14]  Luc Van Gool,et al.  Server-side object recognition and client-side object tracking for mobile augmented reality , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Workshops.

[15]  Heinrich Niemann,et al.  Object recognition using hidden Markov models , 1994 .

[16]  Antonio Torralba,et al.  Context-based vision system for place and object recognition , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[17]  Peter Dayan,et al.  A Neural Substrate of Prediction and Reward , 1997, Science.