TextonBoost: Joint Appearance, Shape and Context Modeling for
Multi-Class Object Recognition and Segmentation. J. Shotton, J. Winn,
C. Rother, A. Criminisi. ECCV 2006.
http://jamie.shotton.org/work/publications/eccv06.pdf
Journal version: http://jamie.shotton.org/work/publications/ijcv07a.pdf
Subscribe to:
Post Comments (Atom)
The authors cite issues with learning the edge potential parameters, \theta_{\phi}, and go on to tune them by hand on the validation set. Has there been any progress made on this?
ReplyDelete