Pith. sign in

REVIEW 4 major objections 7 minor 71 references

Group Downsampling with Equivariant Anti-aliasing

T0 review · 4 major / 7 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Sampling theory for finite groups makes subgroup downsampling perfectly reversible

desk verdict A useful subgroup-sampling routine for G-CNNs whose 'equivariant anti-aliasing' is an unproven approximation rather than an exact operator. read the letter →

arxiv 2504.17258 v1 pith:GEXSMEQI submitted 2025-04-24 cs.LG cs.CVmath.GR

Md Ashiqur Rahman , Raymond A. Yeh This is my paper · ORCID
classification cs.LGcs.CVmath.GR
keywords groupequivariantnetworkssubgroupdownsamplinganti-aliasingsamplingtheoremFouriertransformonfinitegroupsequivariancebandlimitedsignalsCayleygraph
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper aims to show that the familiar CNN operation of downsampling—low-pass filtering followed by keeping every R-th sample—can be generalized from translation grids to feature maps indexed by any finite group. It supplies an algorithm that, given a group and a downsampling rate, returns a suitable subgroup to downsample to, and it defines bandlimited-ness relative to that subgroup. Its central theoretical result, the Subgroup Sampling Theorem, says that a signal whose Fourier coefficients lie in a specific eigenspace can be reconstructed exactly from samples on the subgroup. The paper then constructs an equivariant anti-aliasing projection that approximately enforces that bandlimited condition. If the claim is right, group-equivariant networks can shrink in parameter count and memory without sacrificing accuracy or symmetry guarantees.

What carries the argument

The load-bearing object is the matrix M, a map from Fourier coefficients on the subgroup to Fourier coefficients on the whole group, together with the sampling matrix S and the interpolation operator $I = B F_{G_{\downarrow}}$ where $B = F_G^{-1} M$. The perfect reconstruction condition is that $\hat{x}$ equals $M(M^*M)^{-1}M^*\hat{x}$, and the anti-aliasing operator is the projection $P_M = B(B^*B)^{-1}B^*$ onto the column space of B. The paper constructs M by optimizing Eq. (14), whose equivariance term penalizes deviation from the Reynolds-operator fixed space and whose smoothness term penalizes high graph-Laplacian energy. Subgroup selection is driven by a Cayley-graph algorithm that replaces a generator $s_d$ with its R-th power $s_d^R$, giving a concrete generalization of 'keep every R-th element' to arbitrary finite groups.

What would settle it

Measure the residual equivariance error of the learned anti-aliasing operator directly by computing the norm of $\mathrm{vec}(P_{M^*})$ minus the Reynolds-projected version of $\mathrm{vec}(P_{M^*})$, and check whether it grows with group order or with network depth; alternatively, propagate a fixed input through a deep G-CNN with and without anti-aliasing and test whether the layerwise equivariance error diverges at higher layers. If the residual error grows sharply on larger groups or deeper networks, the empirical equivariance claim would fail.

Watch

Extended reading notes

Core claim

The central claim is the Subgroup Sampling Theorem: for any signal x on a finite group G, if the Fourier coefficient vector $\hat{x}$ lies in the 1-eigenspace of the projection $M(M^*M)^{-1}M^*$, where M connects the Fourier basis of the subgroup $G_{\downarrow}$ to that of G, then x can be reconstructed perfectly from the subsampled signal on $G_{\downarrow}$. The paper finds such an M by solving a constrained optimization that balances an equivariance objective (membership in the fixed space of the Reynolds operator) against a smoothness objective based on the Cayley-graph Laplacian, with exact perfect reconstruction as a constraint. For cyclic groups, the construction reduces to ideal low-pass filtering followed by ordinary subsampling, which is the paper's argument that it generalizes classical sampling theory. Empirically, inserting this downsampling layer into group-equivariant CNNs reduces parameter count, improves or matches accuracy, and lowers equivariance error relative to subsampling without anti-aliasing.

Load-bearing premise

The anti-aliasing projection is only approximately equivariant: equation (14) relaxes exact equivariance into a penalty term with a fixed balancing weight, and the paper gives no bound on how residual equivariance error accumulates through stacked convolution layers.

Editorial extensions

If this is right

  • Group-equivariant CNNs can downsample by an integer rate R along chosen generators, with a concrete subgroup produced by Algorithm 1 instead of requiring hand-picked subgroups.
  • For any such subgroup, anti-aliasing via $P_M$ makes downsampled feature maps losslessly reconstructible from the subgroup samples, provided the bandlimited condition holds.
  • When the group is cyclic, the method reduces to classical downsampling: an ideal low-pass filter followed by subsampling.
  • In image classification experiments, the layer reduces model size while improving or matching accuracy and lowering equivariance error compared with subsampling without anti-aliasing.
  • The anti-aliasing filter needs to be computed only once per architecture before training, because the optimization in Eq. (14) does not depend on the network weights.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the reconstruction condition is linear, one could in principle compose multiple sequential downsampling stages using the same M machinery; the paper does not derive such a composition rule.
  • The equivariance penalty in Eq. (14) uses a fixed balancing weight, so a natural test is whether the residual equivariance error grows with group order or with the depth of the equivariant network.
  • The subgroup-choice heuristic maximizes the number of generators kept; task-adaptive or learned subgroup selection is an untested alternative that could outperform the fixed heuristic on data with anisotropic symmetry content.
  • The interpolation operator gives a concrete way to reconstruct latent group features, which could be used as a regularizer or as a bridge for skip connections in equivariant architectures with multiple resolutions.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 7 minor

Summary. The paper proposes a method for downsampling feature maps on finite groups inside group equivariant CNNs. Given a finite group G and a subsampling rate R, it provides an algorithm (Algorithm 1) that returns a subgroup G_down by modifying the Cayley graph of G, together with a heuristic for choosing which generator to subsample along. It then defines bandlimited signals on G relative to a matrix M satisfying F^{-1}_{G_down} = S F^{-1}_G M (Eq. 9), states a Subgroup Sampling Theorem (Claim 2) guaranteeing perfect reconstruction from subgroup samples for such bandlimited signals, and proposes to construct an anti-aliasing operator P_{M*} by solving the constrained optimization problem in Eq. (14), which trades off an equivariance penalty against a smoothness penalty. Experiments on rotated MNIST, CIFAR-10, and STL-10 report that the proposed layer reduces parameter count, improves classification accuracy, and lowers measured equivariance error Lequi compared to naive subgroup subsampling.

Significance. The paper addresses a real gap in group equivariant architectures: how to downsample a feature map on a general finite group at a specified rate. The proposed subgroup selection algorithm and the interpolation framework are novel, and the authors ship code and provide numerical validation of the reconstruction claim (Table 1). The experimental results consistently show that the anti-aliasing operator improves accuracy and equivariance over naive subsampling in small-scale settings, which is a useful practical contribution. However, the theoretical core is thin: Claim 2 is essentially a direct consequence of the defining constraint in Eq. (9), and the headline claim that the anti-aliasing operator is 'equivariant' is not actually established because Eq. (14) relaxes exact equivariance to a penalty. If the equivariance residual and its propagation are left unquantified, the central claim of the paper is not supported as stated.

major comments (4)
  1. [§4.2, Eq. (14)] The anti-aliasing operator is not proven to be G-equivariant. The exact equivariance condition vec(P_hat_M) = T vec(P_hat_M) is replaced by a penalty term with λ = 5, so the reported P_{M*} is not guaranteed to be exactly equivariant. The paper provides no bound on the residual ||vec(P_hat_{M*}) - T vec(P_hat_{M*})||_2, nor any analysis connecting this residual to the measured Lequi in Table 2 or to the layer-wise equivariance errors in Figures A1-A8. Since 'equivariant anti-aliasing' is a central contribution and the empirical equivariance improvements are attributed to it, this missing quantification is load-bearing. I request that the authors either enforce exact equivariance (e.g., by projecting the solution onto the Reynolds fixed-point space) or provide a residual bound and a propagation analysis, and that they qualify the term 'equivariant' if only approximate equivariance is achieved.
  2. [§4.2, Claim 2] The Subgroup Sampling Theorem is definitional rather than a substantive sampling-theoretic result. The bandlimited subspace is defined as range(B) with B = F^{-1}_G M, and the interpolation operator is defined as I = B F_{G_down}; Eq. (9), namely F^{-1}_{G_down} = S B, is imposed as a constraint. Under these definitions, I S is the identity on range(B) by construction, so Claim 2 is a verification that the constraint achieves what it was designed to achieve. Table 1 numerically validates this identity but does not test an independent hypothesis. The presentation as a 'sampling theorem' generalizing classical sampling theory overstates the novelty; the paper should clearly acknowledge that the reconstruction guarantee is built into the definition of the bandlimited subspace and the interpolation operator.
  3. [Example 3 and Abstract] The claimed recovery of classical low-pass downsampling on cyclic groups is problematic for real-valued signals. The matrix M in Eq. (12) selects the first N/2 complex Fourier coefficients; for a real signal, this coefficient set is not conjugate-symmetric, so the inverse Fourier transform is not real-valued in general. Moreover, for the concrete case C8 → C4, the classical low-pass subspace (DC, the k=1 cosine/sine pair, and the k=2 cosine) splits the 2D real irrep at k=2, so no exactly G-equivariant real projection can have that subspace as its range. Thus the method can only approximately recover the ideal low-pass operation, and the approximation error is not characterized. The abstract and Section 4.2 should be qualified, or a precise condition should be given under which equality (rather than approximation) holds.
  4. [§4.1 and Algorithm 3] There is a direct contradiction between the heuristic described in the text and the provided pseudocode. Section 4.1 states that the algorithm selects the generator with the maximum order satisfying the constraints of Claim 1, but Algorithm 3, line 11, selects the generator with the minimum order via the condition 'if (index = NULL OR O[j] < O[index])'. This makes the subgroup selection procedure ambiguous and prevents reproducibility of the reported experiments from the paper alone. The text or the pseudocode must be corrected, and the choice used in the experiments should be stated explicitly.
minor comments (7)
  1. [§A7] The paper sets λ = 5 in Eq. (14) but provides no sensitivity analysis and no report of the actual equivariance residual of P_{M*} for the groups used in the experiments. Reporting the residual ||vec(P_hat_{M*}) - T vec(P_hat_{M*})||_2 for C24 and D24 would directly address the main concern about approximate equivariance.
  2. [§4.2, Eq. (14)] The displayed optimization objective is typeset in a garbled way in the provided text (the norm, Diag, and absolute-value expressions are not rendered cleanly). The equation should be typeset carefully so that the equivariance objective and smooth selection objective are legible.
  3. [Example 3] The example mixes the complex Fourier basis (used for the matrix M in Eq. (12)) with the real cosine/sine basis discussed in the preceding sentences. The relationship between these bases and the notion of 'ascending frequency' for real signals should be clarified, especially regarding conjugate symmetry.
  4. [Table 1] The 'Recon. Err.' column uses an 'a/b' format to present errors with and without anti-aliasing, but the caption does not define this notation. Please clarify the format in the caption or table note.
  5. [§A3.3] The proof of Claim 1 contains several typographical errors and garbled expressions, including 'od mod r' (should be o_d mod R) and unreadable products such as 'sRnm d' and 'ski_d'. The proof should be carefully proofread and reformatted.
  6. [§A2.1] The heading 'RESUTL' is a typo; it should be 'RESULTS'.
  7. [Algorithm 2] The pseudocode for Check-Compliance lists 'Generators' in the input but then uses the variable s_d without defining it; the notation should be made consistent with Algorithm 1.

Circularity Check

1 steps flagged · score 6.0 of 10

Claim 2's perfect-reconstruction guarantee is definitional: the bandlimited subspace and the interpolation operator are both defined from the same B, so recovery is an algebraic identity.

  1. self definitional [Section 4.2: Eq. (9), Claim 2, and Eq. (11)]
    "Combining Eq. (8) and the fact that x↓ =F^{-1}_{G↓} \hat x↓, we establish the following relationship between M,S and the Fourier bases: F^{-1}_{G↓}=S(F^{-1}_G M)=SB. ... Consecutively, we define the interpolation matrix as I=BF_{G↓}. ... Claim 2. ... if the Fourier coefficients \hat x are in the 1-eigenspace of \bar M ≜M(M†M)^{-1}M† then it can be reconstructed perfectly from the subsampled signal x↓ on G↓."

    Eq. (9) imposes SB=F^{-1}_{G↓}, and I is then defined as BF_{G↓}. Under that constraint, F_{G↓}SB=I, so for every x=B\hat x_c the proof chain Ix↓=BF_{G↓}SB\hat x_c=B\hat x_c=x is an algebraic identity. The theorem's hypothesis, "\hat x in the 1-eigenspace of M(M†M)^{-1}M†", is shown in the proof to be equivalent to x∈col(B). Thus the perfect-reconstruction guarantee is literally the statement that the interpolation operator inverts the sampling operator on its own range, which is built into the definitions of B and I rather than derived from an independent bandlimitedness notion. Table 1 numerically validates this identity on random signals already projected into col(B); it is a consistency check, not a substantive independent prediction.

full rationale

Claim 2 is internally valid but reduces by construction: the paper defines the bandlimited subspace as the range of B=F_G^{-1}M, constrains M by F^{-1}_{G↓}=SB, and defines the interpolation operator as I=BF_{G↓}. The theorem then proves that signals in that subspace are reconstructed, which is exactly the identity I S B=B following from those definitions. This is the paper's central theoretical claim, so it is a real self-definitional step rather than a minor issue. The remaining contributions are non-circular: the subgroup-selection algorithm (Claim 1) is argued from group-theoretic properties, the optimization in Eq. (14) is an explicit design objective, and the classification experiments compare measured accuracies and equivariance errors against external baselines rather than fitting the reported numbers. The citation to Mouli & Ribeiro for the Reynolds-operator characterization is external prior work, not a self-citation chain. I also note a separate, non-circular rigor gap: Eq. (14) only penalizes the equivariance constraint with λ=5, so the resulting operator is not proven exactly G-equivariant; this is a correctness/evidence concern, not a circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central construction rests on standard group-Fourier machinery, a graph-smoothness assumption, and an unverified equivariance-relaxation step, plus the hyperparameter lambda = 5 and the choice of M* from the optimization. These are the costs of the paper's contribution.

free parameters (3)
  • lambda = 5
    Hyperparameter in Eq. (14) balancing the equivariance penalty and the smoothness objective; set manually with no sensitivity analysis reported.
  • Cayley graph Laplacian L = unweighted Laplacian of Gamma(G,S)
    The smooth-selection objective in Eq. (14) uses the Laplacian of the Cayley graph built from the chosen generating set S; different S or edge weights change the notion of smoothness and hence the bandlimited subspace, and this choice is not justified beyond intuition.
  • Matrix M (interpolation/basis matrix) = optimized via SLSQP
    M* is the solution of Eq. (14) and is not unique; the paper's bandlimited subspace depends on the chosen solution. This is a design choice rather than a constant fitted to labels, but it is a degree of freedom in the central construction.
assumptions (5)
  • standard math Peter-Weyl theorem: the set of functions sqrt(d_phi) phi_i^{mn} is an orthonormal basis for functions on G
    Used to define the group Fourier transform and inverse in Eqs. (4)-(5), and to write the signal as x = F_G^{-1} x-hat with an orthonormal basis.
  • standard math The 1-eigenspace of the Reynolds operator characterizes G-equivariant linear maps
    Invoked in section 4.2 (Eqs. 15-17) to express the equivariance constraint as vec(P-hat) = T-bar vec(P-hat), citing Mouli and Ribeiro (2021).
  • domain assumption Cayley graph Laplacian quadratic form measures smoothness of functions on G
    Used in the smooth-selection objective in Eq. (14); the paper adopts a graph-signal-processing notion of smoothness but provides no evidence that this correlates with features useful for classification.
  • ad hoc to paper Relaxing the equivariance constraint to a penalty with lambda = 5 still yields a nearly equivariant filter
    The final anti-aliasing operator P_M* is not exactly equivariant; the architecture's equivariance depends on this unproved approximation, and no error bound is given.
  • domain assumption For the groups used (C24, D24 and similar), Algorithm 1 with the chosen rates produces proper subgroups satisfying Claim 1 conditions
    Claim 1 gives sufficient conditions; the paper verifies these for its experimental groups but does not provide a general characterization of when such a subgroup exists for arbitrary finite groups and rates.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Group Downsampling with Equivariant Anti-aliasing." pith.science (2026). https://pith.science/paper/GEXSMEQI

@misc{pith2026250417258,
  author       = {Pith},
  title        = {Pith review of: Group Downsampling with Equivariant Anti-aliasing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GEXSMEQI}},
  note         = {Machine review of arXiv:2504.17258}
}
read the original abstract

Downsampling layers are crucial building blocks in CNN architectures, which help to increase the receptive field for learning high-level features and reduce the amount of memory/computation in the model. In this work, we study the generalization of the uniform downsampling layer for group equivariant architectures, e.g., G-CNNs. That is, we aim to downsample signals (feature maps) on general finite groups with anti-aliasing. This involves the following: (a) Given a finite group and a downsampling rate, we present an algorithm to form a suitable choice of subgroup. (b) Given a group and a subgroup, we study the notion of bandlimited-ness and propose how to perform anti-aliasing. Notably, our method generalizes the notion of downsampling based on classical sampling theory. When the signal is on a cyclic group, i.e., periodic, our method recovers the standard downsampling of an ideal low-pass filter followed by a subsampling operation. Finally, we conducted experiments on image classification tasks demonstrating that the proposed downsampling operation improves accuracy, better preserves equivariance, and reduces model size when incorporated into G-equivariant networks

Figures

Figures reproduced from arXiv: 2504.17258 by the authors.

Figure 1
Figure 1. (a) Cayley graph of the group with generator ∆t. (b) Edges corresponding to the generator ∆t = 1 are removed (dotted edges) and new edges corresponding to the element 2∆t are added. (c) The resultant cyclic subgroup of size 2 obtain by the traversing the new graph from node 0. Equipped with the intuition, we now introduce two lemmas before going into the conditions and show why Alg. 1 returns G↓ that is a subgroup o… view at source ↗
Figure 2
Figure 2. Subsampling group D8 along the generator s (on left) and r (on right). The edges corresponding to the subsampling generators are dotted in the Cayley graph. Based on this intuition, we propose a heuristic for selecting a set of generators Ds to subsample G with sampling factors Rs along each generator in Ds. Given the subsampling rate R, we decompose it into prime factors, i.e., R = R1 · R2 · R3 · · · , sorted in de… view at source ↗
Figure 3
Figure 3. Visualization of the smoothing filter (PMδG) used in the anti-aliasing operation for subgroup subsampling. The vertical bar corresponds to the value of the filter at each node, with the downward bars indicating negative values [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 64 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    Roto-translation covariant convolutional networks for medical image analysis

    Erik J Bekkers, Maxime W Lafarge, Mitko Veta, Koen AJ Eppenhof, Josien PW Pluim, and Remco Duits. Roto-translation covariant convolutional networks for medical image analysis. In Proc. MICCAI, 2018

    work page 2018

  3. [3]

    A program to build E(N) -equivariant steerable CNNs

    Gabriele Cesa, Leon Lang, and Maurice Weiler. A program to build E(N) -equivariant steerable CNNs . In Proc. ICLR, 2021

    work page 2021

  4. [4]

    Truly shift-invariant convolutional neural networks

    Anadi Chaman and Ivan Dokmanic. Truly shift-invariant convolutional neural networks. In Proc. CVPR, 2021

    work page 2021

  5. [5]

    Sampling theory for graph signals

    Siheng Chen, Aliaksei Sandryhaila, and Jelena Kova c evi \'c . Sampling theory for graph signals. In Proc. ICASSP, 2015 a

    work page 2015

  6. [6]

    Discrete signal processing on graphs: Sampling theory

    Siheng Chen, Rohan Varma, Aliaksei Sandryhaila, and Jelena Kova c evi \'c . Discrete signal processing on graphs: Sampling theory. IEEE TSP, 2015 b

    work page 2015

  7. [7]

    An analysis of single-layer networks in unsupervised feature learning

    Adam Coates, Andrew Ng, and Honglak Lee. An analysis of single-layer networks in unsupervised feature learning. In Proceedings AISTATS. JMLR Workshop and Conference Proceedings, 2011

    work page 2011

  8. [8]

    Group equivariant convolutional networks

    Taco Cohen and Max Welling. Group equivariant convolutional networks. In Proc. ICML, 2016

    work page 2016

Show all 71 references
  1. [9]

    Steerable CNNs

    Taco S Cohen and Max Welling. Steerable CNNs . In Proc. ICLR, 2017

    2017

  2. [10]

    A general theory of equivariant CNNs on homogeneous spaces

    Taco S Cohen, Mario Geiger, and Maurice Weiler. A general theory of equivariant CNNs on homogeneous spaces. In Proc. NeurIPS, 2019

    2019

  3. [11]

    The MNIST database of handwritten digit images for machine learning research

    Li Deng. The MNIST database of handwritten digit images for machine learning research. IEEE SPM, 2012

    2012

  4. [12]

    Groups and the sampling theorem

    MM Dodson. Groups and the sampling theorem. Sampling Theory in Signal and Image Processing, 2007

    2007

  5. [13]

    Learning Laplacian matrix in smooth graph signal representations

    Xiaowen Dong, Dorina Thanou, Pascal Frossard, and Pierre Vandergheynst. Learning Laplacian matrix in smooth graph signal representations. IEEE TSP, 2016

    2016

  6. [14]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at...

    2021

  7. [15]

    A new perspective on building efficient and expressive 3D equivariant graph neural networks

    Yuanqi Du, Limei Wang, Dieqiao Feng, Guifeng Wang, Shuiwang Ji, Carla P Gomes, Zhi-Ming Ma, et al. A new perspective on building efficient and expressive 3D equivariant graph neural networks. In Proc. NeurIPS, 2023

    2023

  8. [16]

    A generalized sampling theorem for locally compact abelian groups

    Adel Faridani. A generalized sampling theorem for locally compact abelian groups. Mathematics of Computation, 1994

    1994

  9. [17]

    A course in abstract harmonic analysis

    Gerald B Folland. A course in abstract harmonic analysis. CRC press, 2016

    2016

  10. [18]

    Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position

    Kunihiko Fukushima. Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position. Biological Cybernetics, 1980

    1980

  11. [19]

    The lie derivative for measuring learned equivariance

    Nate Gruver, Marc Anton Finzi, Micah Goldblum, and Andrew Gordon Wilson. The lie derivative for measuring learned equivariance. In Proc. ICLR, 2023

    2023

  12. [20]

    Deep models of interactions across sets

    Jason Hartford, Devon Graham, Kevin Leyton-Brown, and Siamak Ravanbakhsh. Deep models of interactions across sets. In Proc. ICML, 2018

    2018

  13. [21]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proc. CVPR, 2016

    2016

  14. [22]

    Algebra: a graduate course, volume 100

    I Martin Isaacs. Algebra: a graduate course, volume 100. American Mathematical Soc., 2009

    2009

  15. [23]

    Alias-free generative adversarial networks

    Tero Karras, Miika Aittala, Samuli Laine, Erik H \"a rk \"o nen, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Alias-free generative adversarial networks. In Proc. NeurIPS, 2021

    2021

  16. [24]

    On the generalization of equivariance and convolution in neural networks to the action of compact groups

    Risi Kondor and Shubhendu Trivedi. On the generalization of equivariance and convolution in neural networks to the action of compact groups. In Proc. ICML, 2018

    2018

  17. [25]

    A software package for sequential quadratic programming

    Dieter Kraft. A software package for sequential quadratic programming. Tech. Rep. DFVLR-FB 88-28, 1988

    1988

  18. [26]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky, Geoffrey Hinton, et al. Learning multiple layers of features from tiny images. 2009

    2009

  19. [27]

    Object recognition with gradient-based learning

    Yann LeCun, Patrick Haffner, L \'e on Bottou, and Yoshua Bengio. Object recognition with gradient-based learning. In Shape, contour and grouping in computer vision, 1999

    1999

  20. [28]

    Equiformer: Equivariant graph attention transformer for 3d atomistic graphs

    Yi-Lun Liao and Tess Smidt. Equiformer: Equivariant graph attention transformer for 3d atomistic graphs. In Proc. ICLR, 2023

    2023

  21. [29]

    PIC : permutation invariant critic for multi-agent deep reinforcement learning

    Iou-Jen Liu, Raymond A Yeh, and Alexander G Schwing. PIC : permutation invariant critic for multi-agent deep reinforcement learning. In Proc. CORL, 2020

    2020

  22. [30]

    Semantic tracklets: An object-centric representation for visual multi-agent reinforcement learning

    Iou-Jen Liu, Zhongzheng Ren, Raymond A Yeh, and Alexander G Schwing. Semantic tracklets: An object-centric representation for visual multi-agent reinforcement learning. In Proc. IROS, 2021 a

    2021

  23. [31]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proc. ICCV, 2021 b

    2021

  24. [32]

    A ConvNet for the 2020s

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, and Saining Xie. A ConvNet for the 2020s. In Proc. CVPR, 2022

    2022

  25. [33]

    Invariant and equivariant graph networks

    Haggai Maron, Heli Ben-Hamu, Nadav Shamir, and Yaron Lipman. Invariant and equivariant graph networks. In Proc. ICLR, 2019

    2019

  26. [34]

    A novel sampling theorem on the rotation group

    Jason D McEwen, Martin B \"u ttner, Boris Leistedt, Hiranya V Peiris, and Yves Wiaux. A novel sampling theorem on the rotation group. IEEE SPL, 2015

    2015

  27. [35]

    S peq N ets: Sparsity-aware permutation-equivariant graph networks

    Christopher Morris, Gaurav Rattan, Sandra Kiefer, and Siamak Ravanbakhsh. S peq N ets: Sparsity-aware permutation-equivariant graph networks. In Proc. ICML, 2022

    2022

  28. [36]

    Neural networks for learning counterfactual G -invariances from single environments

    S Chandra Mouli and Bruno Ribeiro. Neural networks for learning counterfactual G -invariances from single environments. In Proc. ICLR, 2021

    2021

  29. [37]

    Geometric invariant theory

    David Mumford, John Fogarty, and Frances Kirwan. Geometric invariant theory. Springer Science & Business Media, 1994

    1994

  30. [38]

    Cyclic spectral analysis of continuous-phase modulated signals

    Antonio Napolitano and Chad M Spooner. Cyclic spectral analysis of continuous-phase modulated signals. IEEE Transactions on Signal Processing, 2001

    2001

  31. [39]

    The completeness of the primitive representations of a closed continuous group

    Fritz Peter and Hermann Weyl. The completeness of the primitive representations of a closed continuous group. Mathematical Annals, 1927

    1927

  32. [40]

    Truly scale-equivariant deep nets with Fourier layers

    Md Ashiqur Rahman and Raymond A Yeh. Truly scale-equivariant deep nets with Fourier layers. In Proc. NeurIPS, 2024

    2024

  33. [41]

    Yeh, Jean Kossaifi, Kamyar Azizzadenesheli, and Anima Anandkumar

    Md Ashiqur Rahman, Robert Joseph George, Mogab Elleithy, Daniel Leibovici, Zongyi Li, Boris Bonev, Colin White, Julius Berner, Raymond A. Yeh, Jean Kossaifi, Kamyar Azizzadenesheli, and Anima Anandkumar. Pretraining codomain attention neural operators for solving multiphysics ...

    2024

  34. [42]

    Deep learning with sets and point clouds

    Siamak Ravanbakhsh, Jeff Schneider, and Barnabas Poczos. Deep learning with sets and point clouds. In Proc. ICLR workshop, 2017

    2017

  35. [43]

    Learnable polyphase sampling for shift invariant and equivariant convolutional networks

    Renan A Rojas-Gomez, Teck-Yian Lim, Alex Schwing, Minh Do, and Raymond A Yeh. Learnable polyphase sampling for shift invariant and equivariant convolutional networks. In Proc. NeurIPS, 2022

    2022

  36. [44]

    Making vision transformers truly shift-equivariant

    Renan A Rojas-Gomez, Teck-Yian Lim, Minh N Do, and Raymond A Yeh. Making vision transformers truly shift-equivariant. In Proc. CVPR, 2024

    2024

  37. [45]

    Romero and Jean-Baptiste Cordonnier

    David W. Romero and Jean-Baptiste Cordonnier. Group equivariant stand-alone self-attention for vision. In Proc. ICLR, 2021

    2021

  38. [46]

    Communication in the presence of noise

    Claude Elwood Shannon. Communication in the presence of noise. Proc. IRE, 1949

    1949

  39. [47]

    The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains

    David I Shuman, Sunil K Narang, Pascal Frossard, Antonio Ortega, and Pierre Vandergheynst. The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains. IEEE Signal Processing Magazine, 2013

    2013

  40. [48]

    Fourier analysis on finite groups with applications in signal processing and system design

    Radomir S Stankovic, Claudio Moraga, and Jaakko Astola. Fourier analysis on finite groups with applications in signal processing and system design. John Wiley & Sons, 2005

    2005

  41. [49]

    Equivariant transformer networks

    Kai Sheng Tai, Peter Bailis, and Gregory Valiant. Equivariant transformer networks. In Proc. ICML, 2019

    2019

  42. [50]

    Cyclic LTI systems in digital signal processing

    PP Vaidyanathan and Ahmet Kirac. Cyclic LTI systems in digital signal processing. IEEE transactions on signal processing, 1999

    1999

  43. [51]

    Impact of aliasing on generalization in deep convolutional networks

    Cristina Vasconcelos, Hugo Larochelle, Vincent Dumoulin, Rob Romijnders, Nicolas Le Roux, and Ross Goroshin. Impact of aliasing on generalization in deep convolutional networks. In Proc. ICCV, 2021

    2021

  44. [52]

    Foundations of signal processing

    Martin Vetterli, Jelena Kova c evi \'c , and Vivek K Goyal. Foundations of signal processing. Cambridge University Press, 2014

    2014

  45. [53]

    A general theory of correct, incorrect, and extrinsic equivariance

    Dian Wang, Xupeng Zhu, Jung Yeon Park, Mingxi Jia, Guanang Su, Robert Platt, and Robin Walters. A general theory of correct, incorrect, and extrinsic equivariance. In Proc. NeurIPS, 2023

    2023

  46. [54]

    General E(2) -equivariant steerable CNNs

    Maurice Weiler and Gabriele Cesa. General E(2) -equivariant steerable CNNs . In Proc. NeurIPS, 2019

    2019

  47. [55]

    Learning steerable filters for rotation equivariant CNNs

    Maurice Weiler, Fred A Hamprecht, and Martin Storath. Learning steerable filters for rotation equivariant CNNs . In Proc. CVPR, 2018

    2018

  48. [56]

    Deep scale-spaces: Equivariance over scale

    Daniel Worrall and Max Welling. Deep scale-spaces: Equivariance over scale. In Proc. NeurIPS, 2019

    2019

  49. [57]

    CvT : Introducing convolutions to vision transformers

    Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai, Lu Yuan, and Lei Zhang. CvT : Introducing convolutions to vision transformers. In Proc. CVPR, 2021

    2021

  50. [58]

    Group equivariant subsampling

    Jin Xu, Hyunjik Kim, Thomas Rainforth, and Yee Teh. Group equivariant subsampling. In Proc. NeurIPS, 2021

    2021

  51. [59]

    E (2) -equivariant vision transformer

    Renjun Xu, Kaifan Yang, Ke Liu, and Fengxiang He. E (2) -equivariant vision transformer. In Proc. UAI, 2023

    2023

  52. [60]

    Deep nets with subsampling layers unwittingly discard useful activations at test-time

    Chiao-An Yang, Ziwei Liu, and Raymond A Yeh. Deep nets with subsampling layers unwittingly discard useful activations at test-time. In Proc. ECCV, 2024

    2024

  53. [61]

    Chirality nets for human pose regression

    Raymond A Yeh, Yuan-Ting Hu, and Alexander Schwing. Chirality nets for human pose regression. In Proc. NeurIPS, 2019 a

    2019

  54. [62]

    Diverse generation for multi-agent sports games

    Raymond A Yeh, Alexander G Schwing, Jonathan Huang, and Kevin Murphy. Diverse generation for multi-agent sports games. In Proc. CVPR, 2019 b

    2019

  55. [63]

    Equivariance discovery by learned parameter-sharing

    Raymond A Yeh, Yuan-Ting Hu, Mark Hasegawa-Johnson, and Alexander Schwing. Equivariance discovery by learned parameter-sharing. In Proc. AISTATS, 2022

    2022

  56. [64]

    Deep sets

    Manzil Zaheer, Satwik Kottur, Siamak Ravanbakhsh, Barnabas Poczos, Ruslan R Salakhutdinov, and Alexander J Smola. Deep sets. In Proc. NeurIPS, 2017

    2017

  57. [65]

    Dive into deep learning

    Aston Zhang, Zachary C Lipton, Mu Li, and Alexander J Smola. Dive into deep learning. Cambridge University Press, 2023

    2023

  58. [66]

    Making convolutional networks shift-invariant again

    Richard Zhang. Making convolutional networks shift-invariant again. In Proc. ICML, 2019

    2019

  59. [67]

    Frequency pooling: Shift-equivalent and anti-aliasing downsampling

    Zhendong Zhang. Frequency pooling: Shift-equivalent and anti-aliasing downsampling. arXiv preprint arXiv:2109.11839, 2021

    2021 arXiv

  60. [68]

    Delving deeper into anti-aliasing in ConvNets

    Xueyan Zou, Fanyi Xiao, Zhiding Yu, and Yong Jae Lee. Delving deeper into anti-aliasing in ConvNets . In Proc. BMVC, 2020

    2020

  61. [69]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...

  62. [70]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...

  63. [71]

    Yeh and Jean Kossaifi and Kamyar Azizzadenesheli and Anima Anandkumar , booktitle= Proc

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

    2024 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.