Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Master Thesis: Structured learning with random forests

Student: -
Advisor: Prof. Dr. Nassir Navab
Supervised by: Chun-Hao Paul Huang, Dr. Federico Tombari.


Recently there has been a significant progress in human pose estimation from range maps and fully 3d data. Thanks to decision forest, now people can infer correspondences (either sparse or dense) between a surface model and single depth image/visual hulls. However, most of the time, the inferred correspondences do not respect the intrinsic structure of meshes, i.e., edge connectivity.

In the previous project, we investigated how to regularize/correct forest predictions with MRF/CRF as a spost-processing step. In this project, we would like take one step further, investigating how to merge MRF/CRF framework into forests, first in the prediction stage and then eventually in the training stage. That is, we hope that forests consider the mesh connectivity to learn and to predict (as in the right figure). This goes into the direction of structured learning. Analogous to the structure-learning version of SVM (termed structured SVM), we would like to see if one can achieve similar things with regression forests.

considering neighbors when drawing predictions


  • The student should have experience in C++ and/or Matlab. Knowledge in random forest is also helpful.

Literature & Resource

  • S. Nowozin, C. Rother, S. Bagon, T. Sharp, B. Yao, and P. Kohli, Decision tree fields (CVPR 2011) paper webpage (with code)
  • J. Jancsary, S. Nowozin, T. Sharp, and C. Rother, Regression Tree Fields An efficient, non-parametric approach to image labeling problems (CVPR 2012) paper webpage (with code)
  • P. Kontschieder, S. Rota, A. Criminisi, P. Kohli, M. Pelillo, and H. Bischof, Context-Sensitive Decision Forests for Object Detection (NIPS 2012)
  • P. Kontschieder, P. Kohli, J. Shotton, and A. Criminisi, GeoF: Geodesic Forests for Learning Coupled Predictors (CVPR 2013)
  • P. Dollar, and L. Zitnick, Structured Forests for Fast Edge Detection (CVPR 2013) code, opencv code
  • S. Nowozin, C. Lampert, Structured Prediction and Learning in Computer Vision, CVPR`11 tutorial webpage


Title: Structured learning with random forests
Director: Prof. Nassir Navab
Supervisor: Chun-Hao Paul Huang, Federico Tombari
Type: DA/MA/BA
Status: draft
Thesis (optional):  

Edit | Attach | Refresh | Diffs | More | Revision r1.9 - 01 Feb 2017 - 14:19 - FedericoTombari