380 likes | 589 Views
POSE–CUT Simultaneous Segmentation and 3D Pose Estimation of Humans using Dynamic Graph Cuts. Mathieu Bray Pushmeet Kohli Philip H.S. Torr Department of Computing Oxford Brookes University. Objective. Image. Segmentation. Pose Estimate. [Images courtesy: M. Black, L. Sigal].
E N D
POSE–CUTSimultaneous Segmentation and 3D Pose Estimation of Humans using Dynamic Graph Cuts Mathieu Bray Pushmeet Kohli Philip H.S. Torr Department of Computing Oxford Brookes University
Objective Image Segmentation Pose Estimate [Images courtesy: M. Black, L. Sigal]
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
The Image Segmentation Problem Segments Image
Problem – MRF Formulation • Notation • Labelling x over the set of pixels • The observed pixel intensity values y (constitute data D) • Energy E (x) = - log Pr(x|D) + constant • Unary term • Likelihood based on colour • Pairwise terms • Prior • Contrast term • Find best labelling x* = arg min E(x)
MRF for Image Segmentation xi = {segment1, …, segmentk} for instance {obj, bkg} Pairwise Potential ij(xi, xj) xi x(labels) xj Unary Potential i(D|xi) i j D(pixels) Image Plane
Can be solved using graph cuts Unary likelihood Contrast Term Ising Model Maximum a-posteriori (MAP) solution x*= Data (D) Unary likelihood Pair-wise Terms MAP Solution MRF for Image Segmentation
MRF for Image Segmentation Need for a human like segmentation Unary likelihood Contrast Term Uniform Prior Maximum-a-posteriori (MAP) solution x*= Data (D) Unary likelihood Pair-wise Terms MAP Solution
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Shape-Priors and Segmentation OBJ-CUT [Kumar et al., CVPR ’05] • Shape-Prior: Layered Pictorial Structure (LPS) • Learned exemplars for parts of the LPS model • Obtained impressive results = + Spatial Layout (Pairwise Configuration) Layer 1 Layer 2
Shape-Priors and Segmentation OBJ-CUT [Kumar et al., CVPR ’05] • Shape-Prior: Layered Pictorial Structure (LPS) • Learned exemplars for parts of the LPS model • Obtained impressive results Unary likelihood colour Shape-Prior Colour + Shape Image
Problems in using shape priors • Intra-class variability • Need to learn an enormous exemplar set • Infeasible for complex subjects (Humans) • Multiple Aspects? • Inference of pose parameters
Do we really need accurate models? • Interactive Image Segmentation [Boykov & Jolly, ICCV’01] • Rough region cues sufficient • Segmentation boundary can be extracted from edges additional segmentation cues user segmentation cues
Do we really need accurate models? • Interactive Image Segmentation • Rough region cues sufficient • Segmentation boundary can be extracted from edges
Rough Shape Prior - The Stickman Model • 26 degrees of freedom • Can be rendered extremely efficiently • Over-comes problems of learning a huge exemplar set • Gives accurate segmentation results
Pose-specific MRF Formulation (pose parameters) Unary Potential i(xi|) Pairwise Potential ij(xi, xj) xi x(labels) xj Unary Potential i(D|xi) i j D(pixels) Image Plane
Pose-specific MRF Energy to be minimized Pairwise potential Unary term Potts model Shape prior distance transform
Pose-specific MRF Energy to be minimized Pairwise potential Unary term Potts model Shape prior + = Colour likelihood colour+ shape MAP Solution Shape Prior Data (D)
What is the shape prior? Energy to be minimized Pairwise potential Unary term Potts model Shape prior How to find the value of ө?
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Resolving ambiguity using multiple views Pose specific Segmentation Energy
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Solving the Minimization Problem To solve: Let F(ө) = Minimize F(ө) using Powell Minimization Computational Problem: Each evaluation of F(ө) requires a graph cut to be computed. (computationally expensive!!) BUT.. Solution: Usethe dynamic graph cut algorithm [Kohli&Torr, ICCV 2005]
solve SA differences between A and B PB* Simpler problem A and B similar SB Dynamic Graph Cuts PA cheaper operation PB computationally expensive operation
solve xa differences between A and B PB* Simpler problem A and B similar Dynamic Graph Cuts 20 msec xb 400 msec
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Segmentation Results Only Colour Colour + Smoothness Colour + Smoothness + Shape Prior Image [Images courtesy: M. Black, L. Sigal]
Segmentation + Pose inference [Images courtesy: M. Black, L. Sigal]
Segmentation + Pose inference [Images courtesy: Vicon]
Outline • Image Segmentation Problem • Pose-Specific Segmentation • The Pose Inference Problem • Optimization • Results • Conclusion and Future Work
Conclusions • Efficient method for using shape priors for object-specific segmentation • Efficient Inference of pose parameters using dynamic graph cuts • Good segmentation results • Pose inference • Needs further evaluation • Segmentation results could be used for silhouette intersection
Future Work • Use dimensionality reduction to reduce the number of pose parameters. • results in less number of pose parameteres to optimize • would speed up inference • Use of features based on texture • Appearance models for individual part of the articulated model (instead of using a single appearance model).