This document discusses regularization techniques for inverse problems. It begins with an overview of compressed sensing and inverse problems, as well as convex regularization using gauges. It then discusses performance guarantees for regularization methods using dual certificates and L2 stability. Specific examples of regularization gauges are given for various models including sparsity, structured sparsity, low-rank, and anti-sparsity. Conditions for exact recovery using random measurements are provided for sparse vectors and low-rank matrices. The discussion concludes with the concept of a minimal-norm certificate for the dual problem.
1. Low Complexity
Regularization of
Inverse Problems
Gabriel Peyré
Joint works with:
Samuel Vaiter Jalal Fadili
Charles Dossal
Mohammad Golbabaee
VISI
www.numerical-tours.com
N
5. Single Pixel Camera (Rice)
x0
˜
y[i] = hx0 , 'i i
P measures
P/N = 1
N micro-mirrors
P/N = 0.16
P/N = 0.02
6. CS Hardware Model
˜
CS is about designing hardware: input signals f
L2 (R2 ).
Physical hardware resolution limit: target resolution f
2
x0 2 L
˜
array
resolution
CS hardware
x0 2 R
N
micro
mirrors
RN .
y 2 RP
7. CS Hardware Model
˜
CS is about designing hardware: input signals f
L2 (R2 ).
Physical hardware resolution limit: target resolution f
x0 2 L
˜
array
resolution
x0 2 R
N
micro
mirrors
y 2 RP
CS hardware
,
,
...
2
RN .
,
Operator
x0
11. Inverse Problem Regularization
Observations: y = x0 + w 2 RP .
Estimator: x(y) depends only on
observations y
parameter
Example: variational methods
1
x(y) 2 argmin ||y
x||2 + J(x)
x2RN 2
Data fidelity Regularity
12. Inverse Problem Regularization
Observations: y = x0 + w 2 RP .
Estimator: x(y) depends only on
observations y
parameter
Example: variational methods
1
x(y) 2 argmin ||y
x||2 + J(x)
x2RN 2
Data fidelity Regularity
Performance analysis:
! Criteria on (x0 , ||w||, ) to ensure
L2 stability ||x(y) x0 || = O(||w||)
Model stability (e.g. spikes location)
14. Union of Linear Models for Data Processing
Union of models: T 2 T linear spaces.
Synthesis
sparsity:
T
Coe cients x
Image
x
15. Union of Linear Models for Data Processing
Union of models: T 2 T linear spaces.
Synthesis
sparsity:
Structured
sparsity:
T
Coe cients x
Image
x
16. Union of Linear Models for Data Processing
Union of models: T 2 T linear spaces.
Synthesis
sparsity:
Structured
sparsity:
T
Coe cients x
Image
x
D
Analysis
sparsity:
Image x
Gradient D⇤ x
17. Union of Linear Models for Data Processing
Union of models: T 2 T linear spaces.
Synthesis
sparsity:
Structured
sparsity:
T
Coe cients x
Analysis
sparsity:
Image
x
D
Low-rank:
Image x
Gradient D⇤ x
S1,·
Multi-spectral imaging:
Pr
xi,· = j=1 Ai,j Sj,·
S2,·
x
S3,·
18. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
+
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
19. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
+
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
Piecewise regular ball , Union of linear models (T )T 2T
x
J(x) = ||x||1 T
T = sparse
vectors
20. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
+
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
Piecewise regular ball , Union of linear models (T )T 2T
T0
x0
x
J(x) = ||x||1 T
T = sparse
vectors
21. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
+
Piecewise regular ball , Union of linear models (T )T 2T
T
T0
x0
x
J(x) = ||x||1 T
T = sparse
vectors
x
T0
x
|x1 |+||x2,3 ||
T = block
sparse
vectors
0
22. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
+
Piecewise regular ball , Union of linear models (T )T 2T
T
T0
x0
x
J(x) = ||x||1 T
T = sparse
vectors
x
T
x
|x1 |+||x2,3 ||
T = block
sparse
vectors
0
x
0
J(x) = ||x||⇤
T = low-rank
matrices
23. Gauges for Union of Linear Models
Gauge:
J :R
N
!R
Convex
8 ↵ 2 R+ , J(↵x) = ↵J(x)
+
Piecewise regular ball , Union of linear models (T )T 2T
T
T0
x0
x
J(x) = ||x||1 T
T = sparse
vectors
T0
x
T
x
|x1 |+||x2,3 ||
T = block
sparse
vectors
0
x
x
x0
0
J(x) = ||x||⇤
T = low-rank
matrices
J(x) = ||x||1
T = antisparse
vectors
29. Examples
`1 sparsity: J(x) = ||x||1
ex = sign(x)
Tx = {z supp(z) ⇢ supp(x)}
P
Structured sparsity: J(x) = b ||xb ||
N (a) = a/||a||
ex = (N (xb ))b2B
Tx = {z supp(z) ⇢ supp(x)}
x
@J(x)
x
0
x
@J(x)
x
0
30. Examples
`1 sparsity: J(x) = ||x||1
ex = sign(x)
Tx = {z supp(z) ⇢ supp(x)}
P
Structured sparsity: J(x) = b ||xb ||
N (a) = a/||a||
ex = (N (xb ))b2B
Tx = {z supp(z) ⇢ supp(x)}
Nuclear norm: J(x) = ||x||⇤
ex = U V
x
@J(x)
x
0
⇤
x
x = U ⇤V ⇤
SVD:
⇤
Tx = {z U? zV? = 0}
@J(x)
x
0
x
@J(x)
31. Examples
`1 sparsity: J(x) = ||x||1
ex = sign(x)
Tx = {z supp(z) ⇢ supp(x)}
P
Structured sparsity: J(x) = b ||xb ||
N (a) = a/||a||
ex = (N (xb ))b2B
Tx = {z supp(z) ⇢ supp(x)}
Nuclear norm: J(x) = ||x||⇤
ex = U V
⇤
x = U ⇤V ⇤
SVD:
⇤
Tx = {z U? zV? = 0}
I = {i |xi | = ||x||1 }
Anti-sparsity: J(x) = ||x||1
Tx = {y yI / sign(xI )}
ex = |I| 1 sign(x)
x
@J(x)
x
0
x
@J(x)
@J(x)
x
0
x
@J(x)
x
x
0
39. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
Sparse vectors: J = || · ||1 .
Theorem: Let s = ||x0 ||0 . If
i,j
⇠ N (0, 1), i.i.d.
[Rudelson, Vershynin 2006]
[Chandrasekaran et al. 2011]
P > 2s log (N/s)
¯
Then 9⌘ 2 D(x0 ) with high probability on
.
40. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
Sparse vectors: J = || · ||1 .
Theorem: Let s = ||x0 ||0 . If
i,j
⇠ N (0, 1), i.i.d.
[Rudelson, Vershynin 2006]
[Chandrasekaran et al. 2011]
P > 2s log (N/s)
¯
Then 9⌘ 2 D(x0 ) with high probability on
Low-rank matrices: J = || · ||⇤ .
Theorem: Let r = rank(x0 ). If
.
[Chandrasekaran et al. 2011]
x0 2 RN1 ⇥N2
P > 3r(N1 + N2 r)
¯
Then 9⌘ 2 D(x0 ) with high probability on
.
41. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
Sparse vectors: J = || · ||1 .
Theorem: Let s = ||x0 ||0 . If
i,j
⇠ N (0, 1), i.i.d.
[Rudelson, Vershynin 2006]
[Chandrasekaran et al. 2011]
P > 2s log (N/s)
¯
Then 9⌘ 2 D(x0 ) with high probability on
Low-rank matrices: J = || · ||⇤ .
Theorem: Let r = rank(x0 ). If
.
[Chandrasekaran et al. 2011]
x0 2 RN1 ⇥N2
P > 3r(N1 + N2 r)
¯
Then 9⌘ 2 D(x0 ) with high probability on
! Similar results for || · ||1,2 , || · ||1 .
.
43. Minimal-norm Certificate
⌘ 2 D(x0 )
=)
⇢
⌘ = ⇤q
ProjT (⌘) = e
Minimal-norm pre-certificate: ⌘0 =
T = T x0
e = ex0
argmin
⌘=
⇤ q,⌘
T =e
||q||
44. Minimal-norm Certificate
⌘ 2 D(x0 )
=)
⇢
⌘ = ⇤q
ProjT (⌘) = e
Minimal-norm pre-certificate: ⌘0 =
Proposition:
One has
T = T x0
e = ex0
argmin
⌘=
⌘0 = (
⇤ q,⌘
+
T
T =e
)⇤ e
||q||
45. Minimal-norm Certificate
⌘ 2 D(x0 )
=)
⇢
⌘ = ⇤q
ProjT (⌘) = e
Minimal-norm pre-certificate: ⌘0 =
argmin
⌘=
One has
⌘0 = (
¯
If ⌘0 2 D(x0 ) and
⇤ q,⌘
+
T
T =e
||q||
)⇤ e
⇠ ||w||,
Proposition:
Theorem:
T = T x0
e = ex0
the unique solution x? of P (y) for y = x0 + w satisfies
Tx ? = T x 0
and ||x?
x0 || = O(||w||) [Vaiter et al. 2013]
46. Minimal-norm Certificate
⌘ 2 D(x0 )
=)
⇢
⌘ = ⇤q
ProjT (⌘) = e
Minimal-norm pre-certificate: ⌘0 =
argmin
⌘=
One has
⌘0 = (
¯
If ⌘0 2 D(x0 ) and
⇤ q,⌘
+
T
T =e
||q||
)⇤ e
⇠ ||w||,
Proposition:
Theorem:
T = T x0
e = ex0
the unique solution x? of P (y) for y = x0 + w satisfies
Tx ? = T x 0
and ||x?
x0 || = O(||w||) [Vaiter et al. 2013]
[Fuchs 2004]: J = || · ||1 .
[Vaiter et al. 2011]: J = ||D⇤ · ||1 .
[Bach 2008]: J = || · ||1,2 and J = || · ||⇤ .
47. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
Sparse vectors: J = || · ||1 .
Theorem: Let s = ||x0 ||0 . If
i,j
⇠ N (0, 1), i.i.d.
[Wainwright 2009]
[Dossal et al. 2011]
P > 2s log(N )
¯
Then ⌘0 2 D( x0 ) wi th hi g h prob a b i l i ty on
.
48. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
i,j
Sparse vectors: J = || · ||1 .
⇠ N (0, 1), i.i.d.
[Wainwright 2009]
[Dossal et al. 2011]
Theorem: Let s = ||x0 ||0 . If
P > 2s log(N )
¯
Then ⌘0 2 D( x0 ) wi th hi g h prob a b i l i ty on
Phase
transitions:
L2 stability
P ⇠ 2s log(N/s)
vs.
.
Model stability
P ⇠ 2s log(N )
49. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
i,j
Sparse vectors: J = || · ||1 .
⇠ N (0, 1), i.i.d.
[Wainwright 2009]
[Dossal et al. 2011]
Theorem: Let s = ||x0 ||0 . If
P > 2s log(N )
¯
Then ⌘0 2 D( x0 ) wi th hi g h prob a b i l i ty on
Phase
transitions:
L2 stability
P ⇠ 2s log(N/s)
vs.
.
Model stability
! Similar results for || · ||1,2 , || · ||⇤ , || · ||1 .
P ⇠ 2s log(N )
50. Compressed Sensing Setting
Random matrix:
2 RP ⇥N ,
i,j
Sparse vectors: J = || · ||1 .
⇠ N (0, 1), i.i.d.
[Wainwright 2009]
[Dossal et al. 2011]
Theorem: Let s = ||x0 ||0 . If
P > 2s log(N )
¯
Then ⌘0 2 D( x0 ) wi th hi g h prob a b i l i ty on
Phase
transitions:
L2 stability
P ⇠ 2s log(N/s)
vs.
.
Model stability
! Similar results for || · ||1,2 , || · ||⇤ , || · ||1 .
P ⇠ 2s log(N )
! Not using RIP technics (non-uniform result on x0 ).