Fredrik Kahl Chalmers University of Technology Collaborators Carl - - PowerPoint PPT Presentation

fredrik kahl
SMART_READER_LITE
LIVE PREVIEW

Fredrik Kahl Chalmers University of Technology Collaborators Carl - - PowerPoint PPT Presentation

Rotation Averaging and Strong Duality Fredrik Kahl Chalmers University of Technology Collaborators Carl Olsson Anders Eriksson Viktor Larsson Tat-Jun Chin Chalmers/Lund University of ETH Zurich University of


slide-1
SLIDE 1

Rotation Averaging and Strong Duality Fredrik Kahl

Chalmers University of Technology

slide-2
SLIDE 2

Collaborators

Carl Olsson Anders Eriksson Viktor Larsson Tat-Jun Chin Chalmers/Lund University of ETH Zurich University of Queensland Adelaide

slide-3
SLIDE 3

Structure from Motion Visual Localization Visual Navigation

slide-4
SLIDE 4

Outline

Main topic: Semidefinite relaxations for optimization over SO(3)

  • Introduction
  • Problem formulation and examples
  • Analysis: Relaxations, tightness and extreme points
  • In depth: Rotation averaging
  • Conclusions
slide-5
SLIDE 5

1 6 5 4 3 2

Rotation averaging

  • Goal: Recover camera poses given relative pairwise measurements
slide-6
SLIDE 6

Hand-eye calibration

slide-7
SLIDE 7

The Chordal distance

  • Defined as the Euclidean distance in the embedding space,
  • Equivalent to:
slide-8
SLIDE 8

Registration of points, lines and planes

slide-9
SLIDE 9

Problem formulation

Let where each .

slide-10
SLIDE 10

How to overcome the problem of non-convexity?

  • One idea: Relax some constraints and solve relaxed problem
  • How to relax?
  • 1. Linearize
  • 2. Convexify
  • Tightness: When is the solution to the original and relaxed problem the same?
slide-11
SLIDE 11

Linearization

  • Longuet-Higgins, 1981
  • Stefanovic, 1973
  • Thompson, 1959
  • Chasles, 1855
  • Hesse, 1863
  • Hauck, 1883
slide-12
SLIDE 12

Convexification

  • Quasi-convexity
  • Semidefinite relaxations
  • F. Kahl, R. Hartley, PAMI 2008
  • Q. Ke, T. Kanade, PAMI 2007
  • F. Kahl, D. Henrion, IJCV 2007
  • C. Aholt, S. Agarwal, R. Thomas, ECCV 2012
slide-13
SLIDE 13

Estimating a single rotation

slide-14
SLIDE 14

Estimating a single rotation

  • Is the relaxation always tight?
  • Are all minimizers Λ* of the convex relaxation rank one?

Original problem Relaxed problem

slide-15
SLIDE 15

Empirical result for 1000 random Q:s

slide-16
SLIDE 16
slide-17
SLIDE 17

Sums of squares polynomials

Multi-variate polynomial p(r) is a sums of squares (SOS) if

slide-18
SLIDE 18

The SO(3)-variety

slide-19
SLIDE 19

A theorem by Blekherman et al, J. Amer. Math. Soc., 2016

slide-20
SLIDE 20

Extreme points

  • Are all minimizers Λ* of the convex relaxation rank one?

Original problem Relaxed problem

slide-21
SLIDE 21

Empirical result for 1000 random Q:s for SO(3)xSO(3)

slide-22
SLIDE 22

Rotation averaging in Structure from Motion

slide-23
SLIDE 23

A possible pipeline:

1.

Estimate relative epipolar geometries (5-point algorithm)

2.

Given relative rotations, estimate absolute rotations

3.

Compute camera positions and 3D points (L -optimization)

Estimate camera poses

slide-24
SLIDE 24

1 6 5 4 3 2

Rotation averaging

  • Goal: Recover camera poses given relative pairwise measurements
slide-25
SLIDE 25

Literature

  • Quaternions:
  • Single rotation estimation:
  • Duality:
  • Analysis:

V.M. Govindu, CVPR 2001 R.I. Hartley, J. Trumpf, Y. Dai and H. Li, IJCV 2013

  • A. Singer, Applied and Computational Harmonic Analysis, 2011
  • J. Fredriksson, C. Olsson, ACCV 2012
  • L. Carlone, D.M. Rosen, G. Calafiore, J.J. Leonard, F. Dellaert,

IROS 2015

  • K. Wilson, D. Bindel and N. Snavely, ECCV 2016
slide-26
SLIDE 26

Rotation averaging

  • Problem formulation

Graph (V,E) where V = camera poses and E = relative rotations 1 6 5 4 3 2

slide-27
SLIDE 27

Rotation averaging

  • Problem formulation

Graph (V,E) where V = camera poses and E = relative rotations

slide-28
SLIDE 28
  • Non-convex problem
  • Three local minima

Rotation averaging

Ground truth Local minimum

slide-29
SLIDE 29
  • Background
  • Well established theory on duality for convex optimization
  • Duality is at the core of many existing optimization algorithms
  • Less understood about the non-convex case
  • Aims
  • Can we obtain guarantees of global optimality?
  • How to design efficient optimization algorithms?

Optimization

slide-30
SLIDE 30

Duality

slide-31
SLIDE 31

Duality

  • Lagrangian:
  • Dual function:
slide-32
SLIDE 32

Duality

Primal problem Dual problem (P) (D) Since (D) is a relaxation of (P), we have

slide-33
SLIDE 33

Primal and dual rotation averaging

Dual problem (P) (D) Lagrangian Primal problem

slide-34
SLIDE 34

Concurrent work

  • D. Cifuentes, S. Agarwal, P. Parrilo, R. Thomas,

”On the Local Stability of Semidefinite Relaxations”, Arxiv 2017

slide-35
SLIDE 35

Main Result Note : Any local minimizer that fulfills this error bound will be global!

slide-36
SLIDE 36

Example:

Corollaries

slide-37
SLIDE 37

Corollaries

Example: For complete graphs,

slide-38
SLIDE 38

Experiments

slide-39
SLIDE 39

Further results

  • Full analysis with proofs
  • New primal-dual algorithm
  • More experimental results
  • A. Eriksson, C. Olsson, F. Kahl, T.J. Chin, to appear PAMI 2019
slide-40
SLIDE 40

Conclusions

  • Strong duality (= zero duality gap) for rotatation averaging provided

bounded noise levels

  • Practically useful sufficient condition for global optimality
  • Analysis also leads to efficient algorithm
slide-41
SLIDE 41

Future work

  • Robust cost functions, e.g., L1 with IRLS
  • Further analysis – when is duality gap zero and for what problems?
slide-42
SLIDE 42

Point averaging

slide-43
SLIDE 43

High-quality night-time images Seasonal changes, urban; Low-quality night-time images Seasonal changes, (sub)urban

Visual localization

www.visuallocalization.net Benchmark challenge and workshop at CVPR 2019

slide-44
SLIDE 44

Estimating a single rotation

  • J. Briales and J. Gonzalez-Jimenez, CVPR 2017
slide-45
SLIDE 45
  • A. Eriksson, C. Olsson, F. Kahl, T.-J. Chin

Rotation Averaging and Strong Duality, CVPR 2018

something else here