Hyperdynamics Importance Sampling

  • Cristian Sminchisescu
  • Bill Triggs
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2350)


Sequential random sampling (‘Markov Chain Monte-Carlo’) is a popular strategy for many vision problems involving multimodal distributions over high-dimensional parameter spaces. It applies both to importance sampling (where one wants to sample points according to their ‘importance’ for some calculation, but otherwise fairly) and to global optimization (where one wants to find good minima, or at least good starting points for local minimization, regardless of fairness). Unfortunately, most sequential samplers are very prone to becoming ‘trapped’ for long periods in unrepresentative local minima, which leads to biased or highly variable estimates. We present a general strategy for reducing MCMC trapping that generalizes Voter’s ‘hyperdynamic sampling’ from computational chemistry. The local gradient and curvature of the input distribution are used to construct an adaptive importance sampler that focuses samples on low cost negative curvature regions likely to contain ‘transition states’ — codimension-1 saddle points representing ‘mountain passes’ connecting adjacent cost basins. This substantially accelerates inter-basin transition rates while still preserving correct relative transition probabilities. Experimental tests on the difficult problem of 3D articulated human pose estimation from monocular images show significantly enhanced minimum exploration.


Hyperdynamics Markov-chain Monte Carlo importance sampling global optimization human tracking 


  1. [1]
    A. Barr. Global and Local Deformations of Solid Primitives. Computer Graphics, 18:21–30, 1984.CrossRefGoogle Scholar
  2. [2]
    M. Black and A. Rangarajan. On the Unification of Line Processes, Outlier Rejection, and Robust Statistics with Applications in Early Vision. IJCV, 19(1):57–92, July 1996.Google Scholar
  3. [3]
    T. Cham and J. Rehg. A Multiple Hypothesis Approach to Figure Tracking. In CVPR, volume 2, pages 239–245, 1999.Google Scholar
  4. [4]
    K. Choo and D. Fleet. People Tracking Using Hybrid Monte Carlo Filtering. In ICCV, 2001.Google Scholar
  5. [5]
    J. Deutscher, A. Blake, and I. Reid. Articulated Body Motion Capture by Annealed Particle Filtering. In CVPR, 2000.Google Scholar
  6. [6]
    J. Deutscher, B. North, B. Bascle, and A. Blake. Tracking through Singularities and Discontinuities by Random Sampling. In ICCV, pages 1144–1149, 1999.Google Scholar
  7. [7]
    S. Duane, A. D. Kennedy, B. J. Pendleton, and D. Roweth. Hybrid Monte Carlo. Physics Letters B, 195(2):216–222, 1987.CrossRefGoogle Scholar
  8. [8]
    D. Forsyth, J. Haddon, and S. Ioffe. The Joy of Sampling. IJCV, 41:109–134, 2001.zbMATHCrossRefGoogle Scholar
  9. [9]
    D. Gavrila and L. Davis. 3-D Model Based Tracking of Humans in Action:A Multiview Approach. In CVPR, pages 73–80, 1996.Google Scholar
  10. [10]
    T. Heap and D. Hogg. Wormholes in Shape Space: Tracking Through Discontinuities Changes in Shape. In ICCV, pages 334–349, 1998.Google Scholar
  11. [11]
    N. Howe, M. Leventon, and W. Freeman. Bayesian Reconstruction of 3D Human Motion from Single-Camera Video. ANIPS, 1999.Google Scholar
  12. [12]
    O. King and D. Forsyth. How does CONDENSATION Behave with a Finite Number of Samples? In ECCV, pages 695–709, 2000.Google Scholar
  13. [13]
    H. J. Lee and Z. Chen. Determination of 3D Human Body Postures from a Single View. CVGIP, 30:148–168, 1985.MathSciNetGoogle Scholar
  14. [14]
    J. MacCormick and M. Isard. Partitioned Sampling, Articulated Objects, and Interface-Quality Hand Tracker. In ECCV, volume 2, pages 3–19, 2000.Google Scholar
  15. [15]
    N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller. Equation of State Calculations by Fast Computing Machines. J. Chem. Phys., 21(6): 1087–1092, 1953.CrossRefGoogle Scholar
  16. [16]
    D. Morris and J. Rehg. Singularity Analysis for Articulated Object Tracking. In CVPR, pages 289–296, 1998.Google Scholar
  17. [17]
    R. Neal. Probabilistic Inference Using Markov Chain Monte Carlo. Technical Report CRG-TR-93-1, University of Toronto, 1993.Google Scholar
  18. [18]
    R. M. Neal. Annealed Importance Sampling. Statistics and Computing, 11:125–139, 2001.CrossRefMathSciNetGoogle Scholar
  19. [19]
    R. Plankers and P. Fua. Articulated Soft Objects for Video-Based Body Modeling. In ICCV, pages 394–401, 2001.Google Scholar
  20. [20]
    R. Rosales and S. Sclaroff. Inferring Body Pose without Tracking Body Parts. In CVPR, pages 721–727, 2000.Google Scholar
  21. [21]
    E. M. Sevick, A. T. Bell, and D. N. Theodorou. A Chain of States Method for Investigating Infrequent Event Processes Occuring in Multistate, Multidimensional Systems. J. Chem. Phys., 98(4), 1993.Google Scholar
  22. [22]
    H. Sidenbladh, M. Black, and D. Fleet. Stochastic Tracking of 3D Human Figures Using 2D Image Motion. In ECCV, 2000.Google Scholar
  23. [23]
    C. Sminchisescu. Consistency and Coupling in Human Model Likelihoods. In CFGR, 2002.Google Scholar
  24. [24]
    C. Sminchisescu and B. Triggs. A Robust Multiple Hypothesis Approach to Monocular Human Motion Tracking. Technical Report RR-4208, INRIA, 2001.Google Scholar
  25. [25]
    C. Sminchisescu and B. Triggs. Covariance-Scaled Sampling for Monocular 3D Body Tracking. In CVPR, 2001.Google Scholar
  26. [26]
    C. Sminchisescu and B. Triggs. Building Roadmaps of Local Minima of Visual Models. In ECCV, 2002.Google Scholar
  27. [27]
    M. R. Sorensen and A. F. Voter. Temperature-Accelerated Dynamics for Simulation of Infrequent Events. J. Chem. Phys., 112(21):9599–9606, 2000.CrossRefGoogle Scholar
  28. [28]
    G. H. Vineyard. Frequency factors and Isotope Effects in Solid State Rate Processes. J. Phys. Chem. Solids, 3:121–127, 1957.CrossRefGoogle Scholar
  29. [29]
    A. F. Voter. A Method for Accelerating the Molecular Dynamics Simulation of Infrequent Events. J. Chem. Phys., 106(11):4665–4677, 1997.CrossRefGoogle Scholar
  30. [30]
    A. F. Voter. Hyperdynamics: Accelerated Molecular Dynamics of Infrequent Events. Physical Review Letters, 78(20):3908–3911, 1997.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Cristian Sminchisescu
    • 1
  • Bill Triggs
    • 1
  1. 1.INRIA Rhône-AlpesMontbonnotFrance

Personalised recommendations