1 / 30

Distribution Fields

Distribution Fields. A Unifying Representation for Low-Level Vision Problems Erik Learned-Miller with Laura Sevilla Lara, Manju Narayana , Ben Mears . Unifying Low-Level Vision. Tracking Optical Flow Affine invariant matching Stereo Backgrounding Registration

amaris
Download Presentation

Distribution Fields

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Distribution Fields A Unifying Representation for Low-Level Vision Problems Erik Learned-Millerwith Laura Sevilla Lara, ManjuNarayana, Ben Mears

  2. Unifying Low-Level Vision • Tracking • Optical Flow • Affine invariant matching • Stereo • Backgrounding • Registration • Image StitchingWhat kind of representation serves all these problems well?

  3. Unifying Descriptors and Representations • HOG, SIFT, Shape Contexts • Geometric blur • Multi-scale methods • Mixture of Gaussian backgrounding • Bilateral filter • Joint alignment (congealing)

  4. Tracking

  5. Basics of Tracking Frame T Frame T+d

  6. Basics of Tracking Frame T Frame T+d

  7. Basics of Tracking Frame T Frame T+d

  8. Basics of Tracking Frame T Frame T+d

  9. The Core Matching Problem Find best match of patch I to image J,for some set of transformations. image J patch I

  10. Key issues • Large “basin of attraction” in gradient descent • Robustness to appearance changes of target

  11. Local optimum problem in alignment Stuck in a local optimum Unaligned

  12. Common solution to gradient descent matching • Blur images? • “Spreads information” • Also destroys information through averaging

  13. Exploding an image One plane for each feature value.

  14. Spatial Blur: 3d convolution with 2d Gaussian

  15. Spatial Blur: 3d convolution with 2d Gaussian KEY PROPERTY: doesn't destroy information through averaging

  16. Properties of Distribution Field Representation • Much larger basin of attraction than other descriptors. • A useful probability model from a single image. • Inherently multi-scale. • Extension and unification of many popular descriptors. • State of the art performance from extremely simple algorithms.

  17. Properties of Distribution Field Representation • Much larger basin of attraction than other descriptors. • A useful probability model from a single image. • Inherently multi-scale. • Extension and unification of many popular descriptors. • State of the art performance with extremely simple algorithms. • THANKS!

  18. The likelihood match

  19. Sharpening match

  20. Understanding the sharpening match What standard deviation maximizes the likelihood of a given point under a zero-mean Gaussian?

  21. Intuition behind sharpening match • Increase standard deviation until it matches “average distance” to matching points.

  22. Properties of the sharpening match • A patch has probability of 1.0 under its own distribution field. • Probability of an image patch degrades gracefully as it is translated away from best position. • Optimum “sigma” value gives a very intuitive notion of the quality of the image match.

  23. Basin of attraction studies

  24. Basin of attraction studies GIVEN A RANDOM PATCH...

  25. Basin of attraction studies AND A RANDOM DISPLACEMENT...

  26. Basin of attraction studies

  27. Basin of attraction studies

  28. Basin of attraction results

  29. Tracking results • State of the art results on tracking with standard sequences • Very simple code • Trivial motion model • Simple memory model

  30. Closely Related work • Mixture of Gaussian backgrounding (Stauffer...) • Shape contexts (Belongie and Malik) • Congealing (me) • Bilateral filter • SIFT (Lowe), HOG (Dalal and Triggs) • Geometric Blur (Berg) • Rectified flow techniques (Efros, Mori) • Mean-shift tracking • Kernel tracking • and many others...

More Related