arxiv: 2605.11589 · v1 · submitted 2026-05-12 · 📡 eess.SP · cs.IT· math.IT

Recognition: 1 theorem link

· Lean Theorem

Unification of Signal Transform Theory

Mitchell A. Thornton

Pith reviewed 2026-05-13 01:18 UTC · model grok-4.3

classification 📡 eess.SP cs.ITmath.IT

keywords signal processingtransform theorygroup representationscovariance analysiseigenbasisunificationPeter-Weyl theoremmatched group

0 comments

The pith

Every listed signal transform is the eigenbasis of covariances invariant under a particular finite or compact group.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes a single representation-theoretic principle that accounts for the discrete Fourier transform, discrete cosine transform, Walsh-Hadamard transform, Haar wavelet transform, Karhunen-Loève transform, and others. Each transform basis is built from the irreducible matrix elements of the group via the Peter-Weyl theorem, so that the basis diagonalizes all covariances unchanged by that group's action. This matters because it turns the choice of transform into the identification of the symmetry group that leaves the data statistics invariant, and it supplies an algorithm to recover that group from an observed covariance matrix. The data-driven Karhunen-Loève transform is recovered as the case with no nontrivial symmetry, while the classical fixed transforms correspond to cyclic, dihedral, abelian, and wreath product groups. The same principle extends to continuous transforms such as the Fourier transform and spherical harmonics.

Core claim

We unify the discrete Fourier transform (DFT), discrete cosine transform (DCT), Walsh-Hadamard, Haar wavelet, Karhunen-Loève transform, and several others along with their continuous counterparts (Fourier transform, Fourier series, spherical harmonics, fractional Fourier transform) under one representation-theoretic principle: each is the eigenbasis of every covariance invariant under a specific finite or compact group, with columns constructed from the irreducible matrix elements of the group via the Peter-Weyl theorem. The unification rests on identifying the matched group of a covariance as the foundational object of second-order signal processing. The data-dependent KLT emerges as the tr

What carries the argument

The matched group of a covariance, the group whose action leaves the covariance invariant and whose irreducible representations yield the transform basis through the Peter-Weyl theorem.

If this is right

The Karhunen-Loève transform corresponds to the trivial matched group case.
Composition of transforms follows from direct, wreath, and semidirect products of their matched groups.
A polynomial-time algorithm exists to discover the matched group of any given covariance by solving a generalized eigenvalue problem.
The fractional Fourier transform is the case for the metaplectic group SO(2) with Hermite-Gauss basis.
The Reed-Muller and arithmetic transforms appear as related change-of-basis transforms on the matched group of Walsh-Hadamard.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Choosing a transform for new data reduces to estimating its invariance group rather than testing multiple bases.
Applications in point cloud processing or 3D vision could automatically select bases matched to the symmetry of their observed covariances.
The inverse relation between matched group size and transform resolution provides a way to balance detail and computation in signal analysis.
Techniques in quantum informatics or single-cell genomics might benefit from discovering natural symmetry groups for their data covariances.

Load-bearing premise

The standard transforms arise exactly as the eigenbases for the covariances invariant under the listed groups, without any post-processing or fitting steps required to match them.

What would settle it

Compute the eigenbasis for a covariance that is exactly invariant under the cyclic group and check whether it matches the discrete Fourier transform basis exactly; any deviation would falsify the claim for that case.

read the original abstract

We unify the discrete Fourier transform (DFT), discrete cosine transform (DCT), Walsh-Hadamard, Haar wavelet, Karhunen-Lo\`eve transform, and several others along with their continuous counterparts (Fourier transform, Fourier series, spherical harmonics, fractional Fourier transform) under one representation-theoretic principle: each is the eigenbasis of every covariance invariant under a specific finite or compact group, with columns constructed from the irreducible matrix elements of the group via the Peter-Weyl theorem. The unification rests on the Algebraic Diversity (AD) framework, which identifies the matched group of a covariance as the foundational object of second-order signal processing. The data-dependent KLT emerges as the trivial-matched-group limit; classical transforms emerge as the cyclic, dihedral, elementary abelian, iterated wreath, and hybrid wreath cases. Composition rules cover direct, wreath, and semidirect products. The Reed-Muller and arithmetic transforms appear as related change-of-basis transforms on the matched group of Walsh-Hadamard. A polynomial-time algorithm for matched-group discovery, the DAD-CAD relaxation cast as a generalized eigenvalue problem in double-commutator form, closes the operational loop: the matched group of any empirical covariance is discovered without expert judgment, with noise-aware variants via the commutativity residual $\delta$ and algebraic coloring index $\alpha$ for finite-SNR settings. The fractional Fourier transform is treated as the metaplectic $SO(2)$ case with Hermite-Gauss matched basis, and a structural principle relates matched group size inversely to transform resolution. Modern applications (massive-MIMO, graph neural networks, transformer attention, point cloud and 3D vision, brain connectivity, single-cell genomics, quantum informatics) are sketched with their matched groups.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The unification claim runs into trouble with non-abelian groups because Peter-Weyl only block-diagonalizes the invariant covariances, not fully diagonalizes them.

read the letter

The paper tries to put DFT, DCT, Walsh-Hadamard, Haar, KLT and a few others under a single representation-theoretic roof by saying each is the eigenbasis of covariances invariant under a particular group, built from irreducible matrix elements via Peter-Weyl. The new pieces are the Algebraic Diversity framework, the matched-group definition, and the DAD-CAD algorithm that turns group discovery into a generalized eigenvalue problem. Those look like genuine additions, and the sketch of how the same idea might apply to massive MIMO or graph networks is at least a reasonable direction to explore. The KLT as the trivial limit case is a clean observation too. The algorithm itself is presented as polynomial-time and noise-aware, which is the sort of operational detail that can make an abstract idea usable. On the positive side, the paper does cite the standard facts about commuting operators and the regular representation, so the basic setup is not invented from scratch. The composition rules for direct, wreath, and semidirect products are also laid out explicitly. The soft spot is exactly where the stress-test note points: for non-abelian groups such as dihedral or iterated wreath products, the higher-dimensional irreps produce blocks of size d_ρ > 1. Peter-Weyl therefore block-diagonalizes every G-invariant covariance, but the eigenvectors inside those blocks are not fixed by the group action alone. You still need an extra choice of basis within each block. The abstract states the transforms are “exactly” the eigenbases without additional fitting or post-hoc adjustments. If the full manuscript does not supply a canonical rule for those intra-block choices or show that they reduce to the classical constructions without extra parameters, the unification does not go through for the non-abelian cases it claims. The abstract gives no derivations or worked examples, so it is impossible to check whether the authors have a fix or have simply overlooked the distinction between block-diagonal and diagonal. This is not a minor technicality; it sits at the center of the strongest claim. The paper is aimed at readers who already work at the intersection of representation theory and signal processing. A serious referee could usefully ask for explicit basis constructions on the dihedral and wreath cases plus a small numerical check that the recovered bases match the known DCT or Walsh matrices. I would send it to review rather than desk-reject, because the algorithmic part and the application sketches are concrete enough to be worth testing, even if the central unification needs substantial repair.

Referee Report

1 major / 2 minor

Summary. The manuscript claims to unify the DFT, DCT, Walsh-Hadamard, Haar, KLT, fractional Fourier transform and their continuous analogues under a single representation-theoretic principle: each transform is the eigenbasis of all covariances invariant under a matched finite or compact group, with its columns given by the irreducible matrix elements via the Peter-Weyl theorem. The Algebraic Diversity (AD) framework is introduced to identify the matched group as the central object of second-order processing; the KLT appears as the trivial-group limit, classical transforms as the cyclic/dihedral/wreath cases, and composition rules are given for direct, wreath and semidirect products. A polynomial-time DAD-CAD relaxation cast as a double-commutator generalized eigenproblem is proposed for data-driven matched-group discovery, together with noise-aware diagnostics and sketches of applications in MIMO, GNNs, transformers and quantum informatics.

Significance. If the unification were rigorously established it would supply a symmetry-based taxonomy for classical transforms and a principled route to new ones, while the matched-group discovery algorithm could become a practical tool for adaptive transform selection. The explicit link to Peter-Weyl and the treatment of the KLT as a limiting case are conceptually attractive and could foster cross-fertilization between signal processing and representation theory.

major comments (1)

[Abstract] Abstract and the statement of the unification principle: the claim that each listed transform (including DCT for the dihedral group and transforms associated with iterated wreath and hybrid wreath products) 'is the eigenbasis of every covariance invariant under a specific finite or compact group, with columns constructed from the irreducible matrix elements of the group via the Peter-Weyl theorem' does not hold for non-abelian groups. In the regular representation, irreps with d_ρ > 1 appear with multiplicity d_ρ; by Schur's lemma every G-invariant operator acts as id_{V_ρ} ⊗ A on the isotypic component, where A is an arbitrary matrix on the multiplicity space. The Peter-Weyl matrix coefficients therefore block-diagonalize all such operators but furnish a common eigenbasis for every invariant covariance only when all irreps are one-dimensional (i.e., the group is abelian). For the dihedr

minor comments (2)

The abstract is overloaded; separating the representation-theoretic unification from the algorithmic contribution and the application sketches would improve readability.
Notation such as 'Algebraic Diversity (AD) framework', 'DAD-CAD relaxation' and 'algebraic coloring index α' is introduced without immediate definition or pointer to the relevant section or prior literature.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the detailed and technically precise report. The observation on the distinction between block-diagonalization and a common eigenbasis for non-abelian groups is well taken and highlights an imprecision in the original abstract. We address this point directly below and will revise the manuscript accordingly.

read point-by-point responses

Referee: [Abstract] Abstract and the statement of the unification principle: the claim that each listed transform (including DCT for the dihedral group and transforms associated with iterated wreath and hybrid wreath products) 'is the eigenbasis of every covariance invariant under a specific finite or compact group, with columns constructed from the irreducible matrix elements of the group via the Peter-Weyl theorem' does not hold for non-abelian groups. In the regular representation, irreps with d_ρ > 1 appear with multiplicity d_ρ; by Schur's lemma every G-invariant operator acts as id_{V_ρ} ⊗ A on the isotypic component, where A is an arbitrary matrix on the multiplicity space. The Peter-Weyl matrix coefficients therefore block-diagonalize all such operators but furnish a common eigenbasis for every invariant covariance only when all irreps are one-dimensional (i.e., the group is abelian). For

Authors: We agree with the referee's analysis. The Peter-Weyl theorem supplies an orthogonal basis of matrix coefficients that reduces every G-invariant operator to block-diagonal form on the isotypic components, with the action on each component of dimension d_ρ² given by I_{d_ρ} ⊗ A. Consequently, the individual columns are common eigenvectors for all invariant covariances if and only if every irrep is one-dimensional. For non-abelian groups the basis is therefore a common reducing basis rather than a common eigenbasis. In the manuscript the DCT is associated with the dihedral group because the conventional covariance class for which the DCT is the eigenbasis (persymmetric Toeplitz matrices arising from even/odd extensions) is a proper subclass of the full dihedral-invariant operators; the additional structure forces A to be diagonal in the chosen basis. The same holds for the wreath-product cases. We will revise the abstract and the statement of the unification principle to read that each transform furnishes the common reducing basis (via irreducible matrix elements) for the covariances invariant under the matched group, with the strict eigenbasis property holding for abelian groups and for the standard non-abelian examples under their conventional additional symmetry constraints. A short paragraph recalling Schur's lemma and the role of multiplicity will be added to the theoretical section. These changes clarify the scope without altering the algorithmic contribution or the listed applications. revision: yes

Circularity Check

1 steps flagged

KLT presented as emerging from framework but definitional by construction of matched-group concept

specific steps

self definitional [Abstract]
"The data-dependent KLT emerges as the trivial-matched-group limit; classical transforms emerge as the cyclic, dihedral, elementary abelian, iterated wreath, and hybrid wreath cases."

The matched-group concept is defined as the group under which the covariance is invariant. For the trivial group, every covariance is invariant by definition, so the corresponding eigenbasis is the KLT by the definition of KLT itself. The claim that it 'emerges as' the limit is therefore tautological within the framework rather than independently derived.

full rationale

The paper's core unification associates each transform with a group such that the transform basis (from Peter-Weyl irreps) diagonalizes all covariances invariant under that group. This association is standard and non-circular for abelian cases like DFT with cyclic groups. However, the explicit statement that the data-dependent KLT 'emerges as the trivial-matched-group limit' reduces directly to the definition: the trivial group leaves all covariances invariant, so its 'eigenbasis' is the KLT by the very definition of KLT as the eigen-decomposition of an arbitrary covariance. No load-bearing self-citations, fitted predictions, or ansatz smuggling are visible in the abstract or described chain; the remainder of the derivation (algorithm for group discovery, composition rules) appears independent of the target result.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The central claim rests on the Peter-Weyl theorem (standard) and the new definitions of Algebraic Diversity and matched group (invented without external validation in the abstract). No free parameters are explicitly fitted in the abstract description.

axioms (1)

standard math Peter-Weyl theorem applies to the finite and compact groups considered
Invoked to construct bases from irreducible matrix elements

invented entities (2)

Algebraic Diversity (AD) framework no independent evidence
purpose: Identifies the matched group of a covariance as the foundational object
New framework introduced to unify transforms
matched group no independent evidence
purpose: Group under which covariance is invariant
Central new concept for second-order signal processing

pith-pipeline@v0.9.0 · 5608 in / 1379 out tokens · 30900 ms · 2026-05-13T01:18:28.045606+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean, IndisputableMonolith/Foundation/AlexanderDuality.lean, IndisputableMonolith/Cost/FunctionalEquation.lean reality_from_one_distinction, J_uniquely_calibrated_via_higher_derivative, alexander_duality_circle_linking unclear
each is the eigenbasis of every covariance invariant under a specific finite or compact group, with columns constructed from the irreducible matrix elements of the group via the Peter-Weyl theorem

Reference graph

Works this paper leans on

58 extracted references · 58 canonical work pages · 5 internal anchors

[1]

Algebraic Diversity: Principles of a Group-Theoretic Approach to Signal Processing

M. A. Thornton, “Algebraic diversity: A group-theoretic frame- work for best-matched-group signal processing,” arXiv preprint arXiv:2604.19983, 2026

work page internal anchor Pith review Pith/arXiv arXiv 2026
[2]

Discrete cosine transform,

N. Ahmed, T. Natarajan, and K. R. Rao, “Discrete cosine transform,”IEEE Transactions on Computers, vol. 23, no. 1, pp. 90–93, 1974

work page 1974
[3]

K. R. Rao and P. Yip,Discrete Cosine Transform: Algorithms, Advantages, Applications. Academic Press, 1990

work page 1990
[4]

Polynomial-Time Optimal Group Selection via the Double-Commutator Eigenvalue Problem

M. A. Thornton, “Polynomial-time optimal group selection via the double-commutator eigenvalue problem,” arXiv preprint arXiv:2605.00834, 2026

work page internal anchor Pith review Pith/arXiv arXiv 2026
[5]

Algebraic Diversity: Group-Theoretic Spectral Estimation from Single Observations

——, “Algebraic diversity: Group-theoretic spectral estimation from single observations,” arXiv preprint arXiv:2604.03634, 2026

work page internal anchor Pith review Pith/arXiv arXiv 2026
[6]

Continuous Algebraic Diversity: Unifying Spectral, Wavelet, and Time-Frequency Analysis via Lie Group Actions

——, “Continuous algebraic diversity: Unifying spectral, wavelet, and time-frequency analysis via Lie group actions,” arXiv preprint arXiv:2605.00848, 2026

work page internal anchor Pith review Pith/arXiv arXiv 2026
[7]

Algebraic signal processing theory: Foundation and 1-D time,

M. P ¨uschel and J. M. F. Moura, “Algebraic signal processing theory: Foundation and 1-D time,”IEEE Transactions on Signal Processing, vol. 56, no. 8, pp. 3572–3585, 2008

work page 2008
[8]

Algebraic signal processing theory: 1-D space,

——, “Algebraic signal processing theory: 1-D space,”IEEE Transactions on Signal Processing, vol. 56, no. 8, pp. 3586– 3599, 2008

work page 2008
[9]

Diaconis,Group Representations in Probability and Statis- tics, ser

P. Diaconis,Group Representations in Probability and Statis- tics, ser. Lecture Notes – Monograph Series. Hayward, CA: Institute of Mathematical Statistics, 1988, vol. 11

work page 1988
[10]

Terras,Fourier Analysis on Finite Groups and Applications, ser

A. Terras,Fourier Analysis on Finite Groups and Applications, ser. London Mathematical Society Student Texts. Cambridge University Press, 1999, vol. 43

work page 1999
[11]

Some applications of generalized FFTs,

D. N. Rockmore, “Some applications of generalized FFTs,” DIMACS Series in Discrete Mathematics and Theoretical Com- puter Science, vol. 28, pp. 329–370, 1995

work page 1995
[12]

Generalized FFTs – a survey of some recent results,

D. K. Maslen and D. N. Rockmore, “Generalized FFTs – a survey of some recent results,”DIMACS Series in Discrete Mathematics and Theoretical Computer Science, vol. 28, pp. 183–237, 1997

work page 1997
[13]

A wreath product group approach to signal and image processing – Part I: Multiresolution analysis,

R. Foote, G. Mirchandani, D. N. Rockmore, D. Healy, and T. Olson, “A wreath product group approach to signal and image processing – Part I: Multiresolution analysis,”IEEE Transactions on Signal Processing, vol. 48, no. 1, pp. 102–132, 2000

work page 2000
[14]

A wreath product group approach to signal and image processing – Part II: Convolution, correlation, and applications,

G. Mirchandani, R. Foote, D. N. Rockmore, D. Healy, and T. Olson, “A wreath product group approach to signal and image processing – Part II: Convolution, correlation, and applications,” IEEE Transactions on Signal Processing, vol. 48, no. 3, pp. 749–767, 2003

work page 2003
[15]

N. S. Jayant and P. Noll,Digital Coding of Waveforms: Prin- ciples and Applications to Speech and Video. Prentice-Hall, 1984

work page 1984
[16]

A theory for multiresolution signal decomposi- tion: the wavelet representation,

S. G. Mallat, “A theory for multiresolution signal decomposi- tion: the wavelet representation,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 11, no. 7, pp. 674–693, 1989

work page 1989
[17]

Vetterli, J

M. Vetterli, J. Kova ˇcevi´c, and V . K. Goyal,Foundations of Signal Processing. Cambridge University Press, 2014

work page 2014
[18]

Theoretical foundations of transform coding,

V . K. Goyal, “Theoretical foundations of transform coding,” IEEE Signal Processing Magazine, vol. 18, no. 5, pp. 9–21, 2001

work page 2001
[19]

Group equivariant convolutional networks,

T. Cohen and M. Welling, “Group equivariant convolutional networks,” inProceedings of the 33rd International Conference on Machine Learning, vol. 48, 2016, pp. 2990–2999

work page 2016
[20]

Geometric deep learning: Going beyond Euclidean data,

M. M. Bronstein, J. Bruna, Y . LeCun, A. Szlam, and P. Van- dergheynst, “Geometric deep learning: Going beyond Euclidean data,”IEEE Signal Processing Magazine, vol. 34, no. 4, pp. 18– 42, 2017

work page 2017
[21]

A practical method for constructing equivariant multilayer perceptrons for arbitrary matrix groups,

M. Finzi, M. Welling, and A. G. Wilson, “A practical method for constructing equivariant multilayer perceptrons for arbitrary matrix groups,” inProceedings of the 38th International Con- ference on Machine Learning, 2021

work page 2021
[22]

Group equivariant stand- alone self-attention for vision,

D. W. Romero and J.-B. Cordonnier, “Group equivariant stand- alone self-attention for vision,” inInternational Conference on Learning Representations, 2021

work page 2021
[23]

Equivariant maps for hierarchical structures,

R. Wang, M. Albooyeh, and S. Ravanbakhsh, “Equivariant maps for hierarchical structures,” inAdvances in Neural Information Processing Systems 33, 2020

work page 2020
[24]

Learning invariances in neural networks from training data,

G. Benton, M. Finzi, P. Izmailov, and A. G. Wilson, “Learning invariances in neural networks from training data,” inAdvances in Neural Information Processing Systems 33, 2020

work page 2020
[25]

Learning disentangled representations and group structure of dynamical environments,

R. Quessard, T. D. Barrett, and W. R. Clements, “Learning disentangled representations and group structure of dynamical environments,” inAdvances in Neural Information Processing Systems 33, 2020

work page 2020
[26]

Generative adversarial symmetry discovery,

J. Yang, R. Walters, N. Dehmamy, and R. Yu, “Generative adversarial symmetry discovery,” inProceedings of the 40th International Conference on Machine Learning, 2023

work page 2023
[27]

On the theory of testing serial correlation,

T. W. Anderson, “On the theory of testing serial correlation,” Scandinavian Actuarial Journal, vol. 1948, no. 3–4, pp. 88–116, 1948

work page 1948
[28]

Regularized estimation of large covariance matrices,

P. J. Bickel and E. Levina, “Regularized estimation of large covariance matrices,”Annals of Statistics, vol. 36, no. 1, pp. 199–227, 2008

work page 2008
[29]

Group symmetry and covari- ance regularization,

P. Shah and V . Chandrasekaran, “Group symmetry and covari- ance regularization,”Electronic Journal of Statistics, vol. 6, pp. 1600–1640, 2012

work page 2012
[30]

Optimal rates of convergence for covariance matrix estimation,

T. T. Cai, C.-H. Zhang, and H. H. Zhou, “Optimal rates of convergence for covariance matrix estimation,”Annals of Statistics, vol. 38, no. 4, pp. 2118–2144, 2010

work page 2010
[31]

A well-conditioned estimator for large-dimensional covariance matrices,

O. Ledoit and M. Wolf, “A well-conditioned estimator for large-dimensional covariance matrices,”Journal of Multivariate Analysis, vol. 88, no. 2, pp. 365–411, 2004

work page 2004
[32]

Serre,Linear Representations of Finite Groups, ser

J.-P. Serre,Linear Representations of Finite Groups, ser. Grad- uate Texts in Mathematics. New York: Springer-Verlag, 1977, vol. 42

work page 1977
[33]

Fulton and J

W. Fulton and J. Harris,Representation Theory: A First Course, ser. Graduate Texts in Mathematics. New York: Springer- Verlag, 1991, vol. 129. 31

work page 1991
[34]

C. W. Curtis and I. Reiner,Representation Theory of Fi- nite Groups and Associative Algebras. New York: Wiley- Interscience, 1962

work page 1962
[35]

R. A. Horn and C. R. Johnson,Matrix Analysis, 2nd ed. Cambridge University Press, 2012

work page 2012
[36]

¨Uber lineare Methoden in der Wahrschein- lichkeitsrechnung,

K. Karhunen, “ ¨Uber lineare Methoden in der Wahrschein- lichkeitsrechnung,”Annales Academiae Scientiarum Fennicae, Series A. I. Mathematica-Physica, vol. 37, pp. 1–79, 1947

work page 1947
[37]

Fonctions al ´eatoires de second ordre,

M. Lo `eve, “Fonctions al ´eatoires de second ordre,”Processus Stochastiques et Mouvement Brownien, 1948

work page 1948
[38]

G. H. Golub and C. F. Van Loan,Matrix Computations, 4th ed. Johns Hopkins University Press, 2013

work page 2013
[39]

Die erste Cohomologiegruppe von ¨uberlagerungen und Homotopie-Eigenschaften dreidimensionaler Mannigfaltigkeiten,

E. Specker, “Die erste Cohomologiegruppe von ¨uberlagerungen und Homotopie-Eigenschaften dreidimensionaler Mannigfaltigkeiten,”Commentarii Mathematici Helvetici, vol. 23, pp. 303–333, 1949

work page 1949
[40]

G. B. Folland,A Course in Abstract Harmonic Analysis, 2nd ed. CRC Press, 2015

work page 2015
[41]

Sugiura,Unitary Representations and Harmonic Analysis: An Introduction, 2nd ed

M. Sugiura,Unitary Representations and Harmonic Analysis: An Introduction, 2nd ed. North-Holland, 1990

work page 1990
[42]

Helgason,Groups and Geometric Analysis: Integral Geome- try, Invariant Differential Operators, and Spherical Functions

S. Helgason,Groups and Geometric Analysis: Integral Geome- try, Invariant Differential Operators, and Spherical Functions. Academic Press, 1984

work page 1984
[43]

N. J. Vilenkin,Special Functions and the Theory of Group Representations, ser. Translations of Mathematical Monographs. American Mathematical Society, 1968, vol. 22

work page 1968
[44]

Die Vollst ¨andigkeit der primitiven Darstellungen einer geschlossenen kontinuierlichen Gruppe,

F. Peter and H. Weyl, “Die Vollst ¨andigkeit der primitiven Darstellungen einer geschlossenen kontinuierlichen Gruppe,” Mathematische Annalen, vol. 97, no. 1, pp. 737–755, 1927

work page 1927
[45]

Functions of positive and negative type, and their connection with the theory of integral equations,

J. Mercer, “Functions of positive and negative type, and their connection with the theory of integral equations,”Philosophical Transactions of the Royal Society of London A, vol. 209, pp. 415–446, 1909

work page 1909
[46]

On unitary representations of the inhomogeneous Lorentz group,

E. Wigner, “On unitary representations of the inhomogeneous Lorentz group,”Annals of Mathematics, vol. 40, no. 1, pp. 149– 204, 1939

work page 1939
[47]

Atkinson and W

K. Atkinson and W. Han,Spherical Harmonics and Approxi- mations on the Unit Sphere: An Introduction, ser. Lecture Notes in Mathematics. Springer, 2012, vol. 2044

work page 2012
[48]

Groemer,Geometric Applications of Fourier Series and Spherical Harmonics

H. Groemer,Geometric Applications of Fourier Series and Spherical Harmonics. Cambridge University Press, 1996

work page 1996
[49]

Differential-difference operators associated to reflection groups,

C. F. Dunkl, “Differential-difference operators associated to reflection groups,”Transactions of the American Mathematical Society, vol. 311, no. 1, pp. 167–183, 1989

work page 1989
[50]

Decomposition of Hardy func- tions into square integrable wavelets of constant shape,

A. Grossmann and J. Morlet, “Decomposition of Hardy func- tions into square integrable wavelets of constant shape,”SIAM Journal on Mathematical Analysis, vol. 15, no. 4, pp. 723–736, 1984

work page 1984
[51]

The emerging field of signal processing on graphs,

D. I. Shuman, S. K. Narang, P. Frossard, A. Ortega, and P. Vandergheynst, “The emerging field of signal processing on graphs,”IEEE Signal Processing Magazine, vol. 30, no. 3, pp. 83–98, 2013

work page 2013
[52]

Discrete signal processing on graphs,

A. Sandryhaila and J. M. F. Moura, “Discrete signal processing on graphs,”IEEE Transactions on Signal Processing, vol. 61, no. 7, pp. 1644–1656, 2013

work page 2013
[53]

Convolu- tional neural networks on graphs with fast localized spectral filtering,

M. Defferrard, X. Bresson, and P. Vandergheynst, “Convolu- tional neural networks on graphs with fast localized spectral filtering,” inAdvances in Neural Information Processing Sys- tems 29, 2016

work page 2016
[54]

Semi-supervised classification with graph convolutional networks,

T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” inInternational Conference on Learning Representations, 2017

work page 2017
[55]

Attention is all you need,

A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” inAdvances in Neural Information Processing Systems 30, 2017

work page 2017
[56]

Quantum Algebraic Diversity: Single-Copy Density Matrix Estimation via Group-Structured Measurements

M. A. Thornton, “Quantum algebraic diversity: Single-copy density matrix estimation via group-structured measurements,” arXiv preprint arXiv:2604.03725, 2026

work page internal anchor Pith review Pith/arXiv arXiv 2026
[57]

M. E. Taylor,Noncommutative Harmonic Analysis, ser. Math- ematical Surveys and Monographs. American Mathematical Society, 1986, vol. 22

work page 1986
[58]

A. W. Knapp,Representation Theory of Semisimple Groups: An Overview Based on Examples. Princeton University Press, 1986

work page 1986