TRINH , MCALLESTER : UNSUPERVISED LEARNING FOR STEREO 1 Unsupervised Learning of Stereo Vision with Monocular Cues

We demonstrate unsupervised learning of a 62 parameter slanted plane stereo vision model involving shape from texture cues. Our approach to unsupervised learning is based on maximizing conditional likelihood. The shift from joint likelihood to conditional likelihood in unsupervised learning is analogous to the shift from Markov random fields (MRFs) to conditional random fields (CRFs). The performance achieved with unsupervised learning is close to that achieved with supervised learning for this model.

[1]  Li Zhang,et al.  Estimating Optimal Parameters for MRF Stereo from a Single Image Pair , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Christopher Joseph Pal,et al.  Learning Conditional Random Fields for Stereo , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Ashutosh Saxena,et al.  3-D Depth Reconstruction from a Single Still Image , 2007, International Journal of Computer Vision.

[4]  Carlo Tomasi,et al.  Multiway cut for stereo and motion with slanted surfaces , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[5]  Pedro F. Felzenszwalb,et al.  Efficiently computing a good segmentation , 1998 .

[6]  Andrea Fusiello,et al.  Quasi-Euclidean uncalibrated epipolar rectification , 2008, 2008 19th International Conference on Pattern Recognition.

[7]  J. M. Hammersley,et al.  Markov fields on finite graphs and lattices , 1971 .

[8]  Jitendra Malik,et al.  Computing Local Surface Orientation and Shape from Texture for Curved Surfaces , 1997, International Journal of Computer Vision.

[9]  David A. McAllester,et al.  Particle Belief Propagation , 2009, AISTATS.

[10]  Andrew Blake,et al.  Shape from Texture: Estimation, Isotropy and Moments , 1990, Artif. Intell..

[11]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[12]  Miguel Á. Carreira-Perpiñán,et al.  On Contrastive Divergence Learning , 2005, AISTATS.

[13]  Andreas Klaus,et al.  Segment-Based Stereo Matching Using Belief Propagation and a Self-Adapting Dissimilarity Measure , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[14]  David A. Forsyth,et al.  Recovering shape and irradiance maps from rich dense texton fields , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[15]  Ashutosh Saxena,et al.  Learning Depth from Single Monocular Images , 2005, NIPS.

[16]  Pedro F. Felzenszwalb,et al.  Efficient belief propagation for early vision , 2004, CVPR 2004.

[17]  J. Aloimonos.  Shape from texture , 1988, Biological cybernetics.

[18]  Ashutosh Saxena,et al.  Depth Estimation Using Monocular and Stereo Cues , 2007, IJCAI.

[19]  A. U.S.,et al.  Recovering Surface Shape and Orientation from Texture , 2002 .

[20]  Hai Tao,et al.  A method for learning matching errors for stereo computation , 2004, BMVC.