1
INSTITUT FÜRNACHRICHTENTECHNIK UNDHOCHFREQUENZTECHNIK
Wavelets and Affine DistributionsA Time-Frequency Perspective
Franz Hlawatsch
Institute of Communications and Radio-Frequency EngineeringVienna University of Technology
– 2 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
Report Documentation Page Form ApprovedOMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.
1. REPORT DATE 07 JAN 2005
2. REPORT TYPE N/A
3. DATES COVERED -
4. TITLE AND SUBTITLE Wavelets and Affine Distributions A Time-Frequency Perspective
5a. CONTRACT NUMBER
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Institute of Communications and Radio-Frequency Engineering ViennaUniversity of Technology
8. PERFORMING ORGANIZATIONREPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited
13. SUPPLEMENTARY NOTES See also ADM001750, Wavelets and Multifractal Analysis (WAMA) Workshop held on 19-31 July 2004.,The original document contains color images.
14. ABSTRACT
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT
UU
18. NUMBEROF PAGES
27
19a. NAME OFRESPONSIBLE PERSON
a. REPORT unclassified
b. ABSTRACT unclassified
c. THIS PAGE unclassified
Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18
2
– 3 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 4 –WAMA-04 Cargèse, France
The notion of time-frequency (TF) analysis
3
– 5 –WAMA-04 Cargèse, France
Auditory perception as TF analysis
– 6 –WAMA-04 Cargèse, France
The TF plane
• Visualize time-frequency location/concentrationof signal x(t):
4
– 7 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 8 –WAMA-04 Cargèse, France
Linear TF analysis
• TF analysis: Measure contribution of TF point to signal
• General approach: Inner product of with “test signal”or “sounding signal” located about :
LTFR = Linear TF Representation
5
– 9 –WAMA-04 Cargèse, France
Linear TF synthesis
• TF synthesis (inversion of LTFR): Recover (“synthesize”) signal from
• General approach:
• Problem: How to construct test (analysis) functions and synthesis functions ?
is represented as superposition of TF localized signal components, weighted by “TF coefficient function”
– 10 –WAMA-04 Cargèse, France
Quadratic TF analysis
• TF analysis: Measure “energy contribution” of TF point to signal
• Simple approach:
• Want QTFR to distribute signal energy over TF plane:
• Problem: How to construct test (analysis) functions ?
QTFR = Quadratic TF Representation
“TF energy distribution”
6
– 11 –WAMA-04 Cargèse, France
Construction of analysis/synthesis functions
• Problem: Construct family of analysis functionssuch that is localized about TF point
• Systematic approach: derived from “prototype function” via unitary “TF displacement operator” :
• Same for synthesis functions :
• Two classical definitions of :– TF shift
– TF scaling (compression/dilatation) + time shift
– 12 –WAMA-04 Cargèse, France
Two classical definitions of operator U
• TF shift:
• TF scaling + time shift:
7
– 13 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 14 –WAMA-04 Cargèse, France
Short-Time Fourier Transform (STFT)
• Recall TF shift:
• ⇒ LTFR = STFT:
STFT = FT of local (windowed) segment of x (t ):
8
– 15 –WAMA-04 Cargèse, France
STFT signal synthesis
• Recall STFT analysis:
• STFT signal synthesis:
is weighted superposition of TF shifted versions of
– 16 –WAMA-04 Cargèse, France
Wavelet Transform (WT)
• Recall TF scaling + time shift:
• ⇒ LTFR = WT:
9
– 17 –WAMA-04 Cargèse, France
WT signal synthesis
• Recall WT analysis:
• WT signal synthesis:
is weighted superposition of TF scaled and time shiftedversions of
– 18 –WAMA-04 Cargèse, France
Spectrogram and scalogram
• Recall LTFR → QTFR:
• STFT → spectrogram:
• WT → scalogram:
10
– 19 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 20 –WAMA-04 Cargèse, France
STFT and constant-BW filterbank: analysis
• STFT analysis as convolution:
• ⇒ Filterbank interpretation/implementation:
11
– 21 –WAMA-04 Cargèse, France
STFT and constant-BW filterbank: synthesis
• STFT synthesis as convolution:
• ⇒ Filterbank interpretation/implementation:
– 22 –WAMA-04 Cargèse, France
Spectrogram analysis as constant-BW filterbank
• Spectrogram analysis as convolution:
• ⇒ Filterbank interpretation/implementation:
12
– 23 –WAMA-04 Cargèse, France
STFT / spectrogram: example
– 24 –WAMA-04 Cargèse, France
WT and constant-Q filterbank: analysis
• WT analysis as convolution:
• ⇒ Filterbank interpretation/implementation:
13
– 25 –WAMA-04 Cargèse, France
WT and constant-Q filterbank: synthesis
• WT synthesis as convolution:
• ⇒ Filterbank interpretation/implementation:
– 26 –WAMA-04 Cargèse, France
Scalogram analysis as constant-Q filterbank
• Scalogram analysis as convolution:
• ⇒ Filterbank interpretation/implementation:
14
– 27 –WAMA-04 Cargèse, France
WT / scalogram: example
– 28 –WAMA-04 Cargèse, France
STFT / spectrogram vs. WT / scalogram
STFT / spectrogram WT / scalogram
15
– 29 –WAMA-04 Cargèse, France
Good-bye and hello
• Good-bye to:
– STFT
– spectrogram
– constant-BW analysis
• Hello to:
– affine class of QTFRs
– Wigner distribution and Bertrand distribution
– hyperbolic TF localization
– 30 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
16
– 31 –WAMA-04 Cargèse, France
Axiomatic (covariance-based) definition of WT
• Generic LTFR expression:
• Covariance of LTFR to TF scalings + time shifts:
• Can show that covariant LTFRs are given by WT
– 32 –WAMA-04 Cargèse, France
Axiomatic (covariance-based) definition of the affine classof QTFRs
• Generic QTFR expression:
• Covariance of QTFR to TF scalings + time shifts:
• Can show that covariant QTFRs are given by
AC = Affine Class
17
– 33 –WAMA-04 Cargèse, France
The affine class of QTFRs
• Affine class of QTFRs:
• 2-D “kernel” specifies QTFR of the AC
• Scalogram is a member of the AC; its kernel is separable:
• Expression of AC QTFRs in terms of signal's FT:
– 34 –WAMA-04 Cargèse, France
Affine class and affine group
• TF scaling + time shift:
• Affine time transformation (“clock change”)
• Composition of clock changes is another clock change:
• ⇒ is unitary representation of the affine group:
– Set:
– Group operation:
– Neutral element:
18
– 35 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 36 –WAMA-04 Cargèse, France
The Wigner-Ville Distribution (WVD)
• Prominent member of the AC: the WVD
• Properties of the WVD:– Covariant to TF scaling and time shift (of course) – Covariant to frequency shift ⇒ not constant-Q
– Real for any (real or complex) signal x(t)
– Marginal properties: e.g.,
– Localization properties: e.g.,for
– Many more…
19
– 37 –WAMA-04 Cargèse, France
Interference terms in the WVD
Interference/cross term
t
ff
t
Source: P. Flandrin, Temps-fréquence. Hermes, Paris, 1993
– 38 –WAMA-04 Cargèse, France
Constant-BW smoothing of the WVD
t
f
Smaller/less interference terms
Poorer TF resolution
Source: P. Flandrin, Temps-fréquence. Hermes, Paris, 1993
20
– 39 –WAMA-04 Cargèse, France
AC expression in terms of WVD
• Any QTFR of the AC can be expressed in terms of the WVD:
where is related to and by FTs
• If is a smooth function, then is a smoothedversion of
• Smoothing causes…– smaller/less interference terms– poorer TF resolution
Affine (constant-Q) smoothing, different from constant-BW smoothing shown on previous slide!
– 40 –WAMA-04 Cargèse, France
Affine (constant-Q) smoothing of the WVD
• Recall:
• Smoothing function at various TF positions:
Smoothing function
21
– 41 –WAMA-04 Cargèse, France
Affine smoothing: example
Smaller/less interference terms
Poorer TF resolution
t
f
Source: P. Flandrin, Temps-fréquence. Hermes, Paris, 1993
– 42 –WAMA-04 Cargèse, France
Scalogram as smoothed WVD
• Recall scalogram:
• Expression of scalogram as smoothed WVD:
Smoothing function is WVD of wavelet:
22
– 43 –WAMA-04 Cargèse, France
Affine WVD smoothing and constant-Q analysis
• Scalogram as smoothed WVD:
– 44 –WAMA-04 Cargèse, France
Constant-BW vs. affine (constant-Q) smoothing
t
f
Smaller/less interference terms
Poorer TF resolution
Source: P. Flandrin, Temps-fréquence. Hermes, Paris, 1993 t
f
t
23
– 45 –WAMA-04 Cargèse, France
OUTLINE
• The notion of time-frequency analysis
• Linear and quadratic time-frequency analysis
• Short-time Fourier transform and wavelet transform;spectrogram and scalogram
• Constant-bandwidth analysis vs. constant-Q analysis
• The affine class
• Affine time-frequency smoothing
• Hyperbolic time-frequency localization
– 46 –WAMA-04 Cargèse, France
Doppler-tolerant signals
• TF scaling / Doppler effect:
• “Doppler-tolerant” signal = eigenfunction of :
• Solution: “hyperbolic impulse”
• Group delay:
Hyperbola in the TF plane
24
– 47 –WAMA-04 Cargèse, France
Example: Bat sonar signals
f
t
HUNTING APPROACH PURSUIT CAPTURE
Source: P. Flandrin
– 48 –WAMA-04 Cargèse, France
Hyperbolic TF localization
• Want AC QTFR to satisfy hyperbolic TF localization property:
• Not satisfied by WVD !
25
– 49 –WAMA-04 Cargèse, France
The Bertrand P0 distribution
• The hyperbolic TF localization property is satisfied by the (unitary) Bertrand P0 distribution
• The Bertrand P0 distribution is a central member of the AC. It satisfies several important properties (besides the hyper-bolic TF localization property).
with
– 50 –WAMA-04 Cargèse, France
Bertrand P0 distribution as generator of the AC
• Any QTFR of the AC can be expressed in terms of the Bertrand P0 distribution:
• Special case: scalogram
where is related to
Smoothing function is BER of wavelet:
26
– 51 –WAMA-04 Cargèse, France
Mellin transform and hyperbolic marginals
• Recall hyperbolic impulse
• Mellin transform:
• Hyperbolic marginal property:
• Not satisfied by WVD… but satisfied by Bertrand P0 distri-bution !
Integrate ACx(t,f) over TF hyperbola t=c/f
– 52 –WAMA-04 Cargèse, France
Application: TF analysis of gravitational wave
Idealized Matched BertrandReassignedspectrogram
WVD Spectrogram Scalogram
Source: E. Chassande-Mottin and P. Flandrin, On the time-frequency detection of chirps. Appl. Comp. Harm. Anal., 6(9): 252-281, 1999.
27
– 53 –WAMA-04 Cargèse, France
Conclusion
• Linear and quadratic TF analysis
• Short-time Fourier transform and spectrogram
• Wavelet transform and scalogram
• Filterbank interpretation: constant-BW analysis versus constant-Q analysis
• Scaling/shift covariance and affine class of QTFRs
• Wigner-Ville distribution and affine smoothing
• Doppler tolerance and hyperbolic impulses
• Hyperbolic TF localization and Bertrand P0 distribution
• Mellin transform and hyperbolic marginal property
– 54 –WAMA-04 Cargèse, France
WARNING
YOU ARE LEAVING THE YOU ARE LEAVING THE TIMETIME--FREQUENCY PLANEFREQUENCY PLANE