Perceptive Kuramoto Oscillators - PeKO Martin Meier September 9, 2013 Martin Meier () PeKO September 9, 2013 1/1
Perceptive Kuramoto Oscillators - PeKO
Martin Meier
September 9, 2013
Martin Meier () PeKO September 9, 2013 1 / 1
Motivation
Synchrony is a natural phenomenon
Martin Meier () PeKO September 9, 2013 2 / 1
Motivation
Outline
Martin Meier () PeKO September 9, 2013 3 / 1
Recap
Recap: Oscillator Network
I Previous talk: Perceptual Grouping with Oscillators
I Oscillator described by phase θ and frequency ω
I Phase update:
θ̇m = ωm +K
N
N∑n=1
Fmnsin(θn − θm)
I Frequency update:
ωm = ω0 · argmaxα
( ∑n∈N (α)
Fmn ·1
2(cos(θn − θm) + 1)
)
Martin Meier () PeKO September 9, 2013 4 / 1
Recap
Recap: Oscillator Network
I Previous talk: Perceptual Grouping with Oscillators
I Oscillator described by phase θ and frequency ω
I Phase update:
θ̇m = ωm +K
N
N∑n=1
Fmnsin(θn − θm)
I Frequency update:
ωm = ω0 · argmaxα
( ∑n∈N (α)
Fmn ·1
2(cos(θn − θm) + 1)
)
Martin Meier () PeKO September 9, 2013 4 / 1
Recap
Recap: Oscillator Network
I Previous talk: Perceptual Grouping with Oscillators
I Oscillator described by phase θ and frequency ω
I Phase update:
θ̇m = ωm +K
N
N∑n=1
Fmnsin(θn − θm)
I Frequency update:
ωm = ω0 · argmaxα
( ∑n∈N (α)
Fmn ·1
2(cos(θn − θm) + 1)
)
Martin Meier () PeKO September 9, 2013 4 / 1
Recap
Recap: Oscillator Network
I Previous talk: Perceptual Grouping with Oscillators
I Oscillator described by phase θ and frequency ω
I Phase update:
θ̇m = ωm +K
N
N∑n=1
Fmnsin(θn − θm)
I Frequency update:
ωm = ω0 · argmaxα
( ∑n∈N (α)
Fmn ·1
2(cos(θn − θm) + 1)
)
Martin Meier () PeKO September 9, 2013 4 / 1
Recap
Recap: Evaluation
I Comparison to the CLM, similar settings for both
I IA matrix with 1000 features in ten groups
I 100 layers, 100 discrete frequencies
I All with different amounts of noise in the IA matrices
I 500 trials for each condition
Martin Meier () PeKO September 9, 2013 5 / 1
Recap
Recap: Evaluation Results
I Previous talk: Evaluation results
I Evaluation revealed:
I Quality comparable to the CLMI Computational complexity reducedI Grouping speed increased
Martin Meier () PeKO September 9, 2013 6 / 1
Recap
Recap: Evaluation Results
I Previous talk: Evaluation resultsI Evaluation revealed:
I Quality comparable to the CLM
I Computational complexity reducedI Grouping speed increased
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25 30 35 40 45
gro
up
ing q
ualit
y, m
ean a
nd
std
dev
% of noise
CLMOscillators
Martin Meier () PeKO September 9, 2013 6 / 1
Recap
Recap: Evaluation Results
I Previous talk: Evaluation resultsI Evaluation revealed:
I Quality comparable to the CLMI Computational complexity reducedI Grouping speed increased
0
50
100
150
200
250
300
0 5 10 15 20 25 30# o
f up
date
ste
ps,
mean a
nd s
tddev
% of noise
CLMOscillators
31 32 33 34 35 36 37 38 39 0
100
200
300
400
500
600
700
800
900
1000
# o
f update
ste
ps,
mean a
nd s
tddev
% of noise
CLMOscillators
Martin Meier () PeKO September 9, 2013 6 / 1
Recap New Evaluation
New: Robustness to Perturbations
I Both models converge for 500 steps
I Split target groups (10 → 20)
I Measure #steps needed for new grouping result
Martin Meier () PeKO September 9, 2013 7 / 1
Recap New Evaluation
New: Robustness to Perturbations
I Both models converge for 500 steps
I Split target groups (10 → 20)
I Measure #steps needed for new grouping result
1
10
100
5 10 15 20 25 30 35
avera
ge n
um
ber
of
steps
% of noise
CLMOscillators
Martin Meier () PeKO September 9, 2013 7 / 1
Oscillation and Order
Stablility
I PeKO achieves good grouping results.
I How to assess the grouping quality?
Martin Meier () PeKO September 9, 2013 8 / 1
Oscillation and Order
Kuramoto Order Parameter
I Complex parameter, r and ψ
I re iψ =1
N
N∑n=1
e iθn
I r is phase coherence ∈ [0, 1]
I ψ is average phase
Re
Im
ψ
R
complex plane
Martin Meier () PeKO September 9, 2013 9 / 1
Oscillation and Order
Order Parameter
I Needs to be adapted for discrete frequencies:
rαeiψα =
1
Nα
Nα∑n=1
e iθn if Nα 6= 0
Martin Meier () PeKO September 9, 2013 10 / 1
Oscillation and Order
Order Behavior
0
0.2
0.4
0.6
0.8
1
0 5 10 15 20 25 30 35 40 45
gro
up
ing q
ualit
y, m
ean a
nd
std
dev
% of noise
CLMOscillators
I Recap: Grouping quality with respect to noise
I Order parameter r̄ wrt. noise and time
I Order parameter can be used to assess grouping quality
Martin Meier () PeKO September 9, 2013 11 / 1
Oscillation and Order
Order Behavior
0
5
10
15
20
25
30
35
40
45
0 200 400 600 800 1000
am
ount
of
nois
e
update step
0
0.2
0.4
0.6
0.8
1
ord
er
0.90.8
0.7
0.6
0.5
0.6
0.5
0.40.3
I Recap: Grouping quality with respect to noise
I Order parameter r̄ wrt. noise and time
I Order parameter can be used to assess grouping quality
Martin Meier () PeKO September 9, 2013 11 / 1
Oscillation and Order
Order Behavior
0
5
10
15
20
25
30
35
40
45
0 200 400 600 800 1000
am
ount
of
nois
e
update step
0
0.2
0.4
0.6
0.8
1
ord
er
0.90.8
0.7
0.6
0.5
0.6
0.5
0.40.3
I Recap: Grouping quality with respect to noise
I Order parameter r̄ wrt. noise and time
I Order parameter can be used to assess grouping quality
Martin Meier () PeKO September 9, 2013 11 / 1
Model Extensions
Dealing with spurious features
I Everyone knows: Not every feature is relevant
I We have to deal with them
(a) IA Matrix. (b) Step 1. (c) Step 2. (d) Step 3. (e) Step 50.
Martin Meier () PeKO September 9, 2013 12 / 1
Model Extensions
Dealing with spurious features
I Everyone knows: Not every feature is relevant
I We have to deal with them
(f) IA Matrix. (g) Step 1. (h) Step 2. (i) Step 3. (j) Step 50.
Martin Meier () PeKO September 9, 2013 12 / 1
Model Extensions
Dealing with spurious features
I Background Layer: Nice idea “borrowed” from the CLM
I Introduced as special frequency
I Possesses “chaotic” coupling:
θ̇m = ωm + Kr sin(ψ − θm).
Martin Meier () PeKO September 9, 2013 13 / 1
Model Extensions
Dealing with spurious features
I Same example as before
I Spurious features are collected by the background frequency
(k)Initialization.
(l) Step 1. (m) Step 2. (n) Step 3. (o) Step 50.
Martin Meier () PeKO September 9, 2013 14 / 1
Model Extensions
Real World Example: Texture Grouping
(p) Input image. (q) Without background. (r) With background.
Martin Meier () PeKO September 9, 2013 15 / 1
Improved Learning
Improved Learning of Lateral Interactions
I Idea from S. Weng
I Represent feature compatibility by distance functions
I Create prototypes with VQ
I Labeled examples are used to decide if +/− interaction
Martin Meier () PeKO September 9, 2013 16 / 1
Improved Learning
Improved Learning of Lateral Interactions
I Original approach used Activity Equilibrium VQI Replaced by ITVQ
I Better distribution of prototypes
I Evaluated in contour grouping task
Martin Meier () PeKO September 9, 2013 17 / 1
Improved Learning
Improved Learning of Lateral Interactions
I Evaluated with three kinds of shapes
I 200 trials for each shapeI Four conditions
I CLM with AEVI PeKO with AEVI CLM with ITVQI PeKO with ITVQ
Martin Meier () PeKO September 9, 2013 18 / 1
Improved Learning
Feature and Data Example
(s) Oriented edge features. (t) “Easy” problem. (u) “Hard” problem.
Martin Meier () PeKO September 9, 2013 19 / 1
Improved Learning
Example of generated Prototypes
C+C-
Reference
C+C-
Reference
ITVQ Prototypes AEV Prototypes
Martin Meier () PeKO September 9, 2013 20 / 1
Improved Learning
Improved Learning of Lateral Interactions - Results
0
0.2
0.4
0.6
0.8
1
GroupingQuality
AEV CLMAEV Oscillators
ITVQ CLMITVQ Oscillators
50 Prototypes 100 Prototypes 150 Prototypes 200 Prototypes
Martin Meier () PeKO September 9, 2013 21 / 1
Conclusion
Conclusion
I Oscillators are robust to perturbations
I Order allows assessment of grouping quality
I “Chaotic” frequency handles spurious features
I Learing of lateral interactions is improved
Martin Meier () PeKO September 9, 2013 22 / 1
Conclusion
Conclusion
I Oscillators are robust to perturbations
I Order allows assessment of grouping quality
I “Chaotic” frequency handles spurious features
I Learing of lateral interactions is improved
Martin Meier () PeKO September 9, 2013 22 / 1
Conclusion
Conclusion
I Oscillators are robust to perturbations
I Order allows assessment of grouping quality
I “Chaotic” frequency handles spurious features
I Learing of lateral interactions is improved
Martin Meier () PeKO September 9, 2013 22 / 1
Conclusion
Conclusion
I Oscillators are robust to perturbations
I Order allows assessment of grouping quality
I “Chaotic” frequency handles spurious features
I Learing of lateral interactions is improved
Martin Meier () PeKO September 9, 2013 22 / 1
Conclusion
Thank You!
Any Questions?
Martin Meier () PeKO September 9, 2013 23 / 1
Conclusion
Thank You!
Any Questions?
Martin Meier () PeKO September 9, 2013 23 / 1
ITVQ Update Rule
I Minimize Cauchy-Schwartz Divergence
I x0 is input, x prototypes
I Fixed point update rule:
x t+1i =
∑N0j=1 Gσ(x ti − x0j)x0j∑N0j=1 Gσ(x ti − x0j)
− c
∑Nj=1 Gσ(x ti − x tj )x tj∑N0j=1 Gσ(x ti − x0j)
+c
∑Nj=1 Gσ(x ti − x tj )∑N0j=1 Gσ(x ti − x0j)
x ti ; c =N0
N
V (X ,X0)
V (X )
Martin Meier () PeKO September 9, 2013 24 / 1
Learning of Lateral Interactions
+
labeled training data
1
f >0
f <0
proximityvector d(v ,v )
Voronoi cellproximity prototype
r r'
rr'
rr'
proximity space D
clustering
positive coefficents
negative coefficents interaction function
2
3
Martin Meier () PeKO September 9, 2013 25 / 1