Pith. sign in

REVIEW 4 major objections 6 minor 48 references

PSDNet and DPDNet: Efficient channel expansion, Depthwise-Pointwise-Depthwise Inverted Bottleneck Block

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper argues that depthwise convolution can replace 1x1 convolution for expanding channel counts in bottleneck blocks, yielding networks with fewer parameters and FLOPs and comparable or better accuracy.

desk verdict A rough draft with a real parameter-efficiency trick, no ImageNet results, and an overreaching mechanism claim; worth a serious look for the small but honest CIFAR evidence. read the letter →

arxiv 1909.01026 v2 pith:JCDKA3RK submitted 2019-09-03 cs.CV

Guoqing Li , Meng Zhang , Qianru Zhang , Ziyang Chen , Wenzhao Liu , Jiaojie Li , Xuzhao Shen , Jianjun Li
show 2 more authors
Zhenyu Zhu Chau Yuen
This is my paper · ORCID
classification cs.CV
keywords depthwiseconvolutionchannelexpansionbottleneckblockinvertedlightweightCNNPSDNetDPDNetparameterefficiency
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

This paper argues that the common practice of using 1x1 pointwise convolutions to expand channels in bottleneck blocks is less efficient than using depthwise convolutions (DWCs) for the same job. It derives a simple ratio showing that, for a kernel size k and C input channels, DWC beats PWC by a factor of C/$k^{2}$ in both parameters and computation. Based on this, it builds PSDNet, a ResNet-style network with DWC expansion, and DPDNet, an inverted-bottleneck network using a depthwise-pointwise-depthwise (DPD) block. On CIFAR-10 and CIFAR-100, PSDNet improves on the compressed ResNet50 baseline while cutting parameters from 2.0M to 1.6M, and DPDNet matches MobileNetV2 accuracy at roughly 60% of its parameters.

What carries the argument

The load-bearing identity is the parameter and cost ratio C/$k^{2}$ between pointwise and depthwise channel expansion (Eq. 1 and Eq. 2): a depthwise convolution with kernel size k and channel multiplier m needs k*k*mC parameters and W*H*k*k*mC operations, versus C*mC for a 1x1 convolution, so the DWC wins whenever the channel count C exceeds $k^{2}$. The second machinery is the DPD block, an inverted bottleneck made of an expanding DWC (m>1), a channel-mixing PWC, and a per-channel DWC (m=1), which the paper stacks to form DPDNet and tunes with width multiplier $\alpha$ and channel multiplier m.

What would settle it

Measure wall-clock latency and energy on a representative mobile or edge accelerator for DPDNet and MobileNetV2 at matched parameter budgets: if DWC-based models are not faster despite fewer FLOPs, the central efficiency claim is false in practice. Alternatively, train MobileNetV2 with the DPD block's layer ordering but identical channel counts and training settings; if the accuracy gap vanishes, the paper's attribution to DWC expansion is not supported.

Watch

Extended reading notes

Core claim

The central discovery is that depthwise convolution, normally used with channel multiplier m=1 so input and output channel counts match, can instead be used with m>1 to expand channels in a bottleneck block. The paper proves that for expanding a feature map from C to mC channels, a k x k DWC has C/$k^{2}$ times fewer parameters and FLOPs than a 1x1 PWC, since the PWC costs C*mC while the DWC costs k*k*mC. Networks built on DWC expansion, PSDNet and DPDNet, report fewer parameters, lower FLOPs, and accuracy equal to or better than their PWC-based counterparts. The authors further report that models with more DWC layers outperform models with more PWC layers, interpreting this as evidence that extracting spatial features matters more than combining channel information.

Load-bearing premise

The argument assumes that parameter count and FLOPs are the right measures of efficiency and that the reported accuracy gaps are caused by the DWC-versus-PWC block ordering rather than by other architectural differences.

Editorial extensions

If this is right

  • PSDNet50 cuts parameters from 2.0M to 1.6M and raises CIFAR-10 accuracy from 92.95% to 93.87% versus the compressed ResNet50 baseline.
  • DPDNet uses about 60% of MobileNetV2's parameters at the same m and alpha, with accuracy within about a percentage point on CIFAR-10/100 across most settings.
  • Two hyperparameters, alpha and m, give a smooth accuracy-versus-cost trade-off, letting one network family cover a range of deployment budgets.
  • The DWC-first ordering in DPD blocks appears to outperform the PWC-first ordering of MobileNetV2 when parameter counts are matched, suggesting spatial feature extraction is more valuable per parameter than channel combination.

Reading between the lines

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

  • The paper's efficiency metric is theoretical FLOPs; depthwise convolutions are memory-bound on many real accelerators, so the reported parameter and FLOP savings may not translate to proportional latency gains without explicit hardware benchmarking.
  • The accuracy comparison with MobileNetV2 is confounded by different block structure and channel counts; a cleaner test would hold block order constant and vary only the expansion layer type.
  • The C/k^2 ratio suggests the advantage of DWC expansion grows with channel count, so the benefit should be largest in the last layers of deep networks; measuring layer-wise sensitivity would test this.
  • If the spatial-feature hypothesis holds, DWC expansion should show larger gains on spatially demanding tasks like object detection or segmentation than on classification.
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 / 6 minor

Summary. The paper proposes replacing the pointwise (1×1) convolution used for channel expansion in bottleneck blocks with a depthwise convolution (DWC) whose channel multiplier m is greater than 1. The authors derive parameter and FLOP ratios in Eqs. (1)–(2) and show that DWC expansion is cheaper than PWC expansion by a factor of C/k². Based on this idea, they introduce PSDNet, a ResNet variant that replaces the final 1×1 expansion layer with a DWC layer, and DPDNet, a lightweight network built from a Depthwise-Pointwise-Depthwise (DPD) inverted bottleneck block. Experiments on CIFAR-10, CIFAR-100, and CINIC-10 report parameter and FLOP reductions with comparable or better accuracy relative to ResNet and MobileNetV2, along with sweeps over the hyperparameters m and α. An ImageNet evaluation is described in the text but not actually reported.

Significance. The arithmetic in Eqs. (1) and (2) is correct, and the parameter and FLOP counts in Tables 4–6 are internally consistent, giving a concrete efficiency comparison for the proposed blocks. The two hyperparameters m and α provide a useful accuracy–cost trade-off, and the CINIC-10 results at matched parameter counts are a genuine empirical contribution. The central proposal—using DWC for channel expansion—is simple and plausibly effective at reducing parameter cost. However, the experiments do not isolate the proposed mechanism: the compared blocks differ in multiple structural ways, accuracy differences are often below 1% with no error bars, and the promised ImageNet results are absent. The significance of the claim that DWC expansion is both cheaper and more accurate than PWC expansion is therefore not yet established.

major comments (4)
  1. [Section 3.1, Eqs. (1)–(2)] The parameter and FLOP comparison treats DWC(m>1) and PWC as equivalent channel-expansion operators, but they are not functionally equivalent. Each output channel of a DWC(m>1) layer is derived from a single input channel with m filters, whereas every PWC output channel mixes all C input channels. Replacing a PWC expansion layer with a DWC therefore changes the information flow of the block, not merely its parameter cost. Consequently, the accuracy differences between PSDNet and ResNet in Table 4 (0.9% on CIFAR-10, 0.5% on CIFAR-100) conflate the efficiency gain with a structural change. To support the central claim, the paper needs an ablation that holds the block structure as constant as possible and varies only the expansion operation, with results averaged over multiple seeds.
  2. [Section 3.3, Tables 5 and 6] The DPDNet-versus-MobileNetV2 comparison changes several architectural factors simultaneously: layer order (DWC-PWC-DWC versus PWC-DWC-PWC), the layer that performs expansion, the number of DWC versus PWC layers, and stride handling. The reported accuracy differences are frequently below 1% and sometimes favor MobileNetV2 (e.g., CIFAR-100 for m=3–6 in Table 5; CIFAR-10 for α=2.5, 3.0, 4.0 in Table 6). Because the blocks differ in more than the expansion operation, these results do not isolate the effect of using DWC for expansion. The paper should provide matched comparisons in which only the expansion mechanism differs, or explicitly control for the other structural variables, and should report multiple runs with error bars.
  3. [Section 3.2.2 and Table 8] The ImageNet evaluation is incomplete: the text contains placeholders such as "M parameters" and "Table x", the DPDNet row in Table 8 is empty, and the conclusion states that evaluation on ImageNet is future work. The abstract and introduction claim that DPDNet is more efficient than MobileNetV2 without dataset qualification. The paper must either provide the ImageNet results or explicitly and prominently scope the claims to the CIFAR and CINIC-10 datasets; as written, the general efficiency claim is unsupported.
  4. [Tables 4–6] All accuracy numbers are reported for a single run, with no error bars, multiple seeds, or statistical significance tests. Many differences are smaller than 1% (e.g., Table 5, CIFAR-10 m=3–5; Table 6, CIFAR-10 α=1.25–2.0), so without variance estimates the "higher accuracy" and "approximate accuracy" claims are not robustly supported. The authors should run each configuration at least three times and report mean and standard deviation.
minor comments (6)
  1. [Section 4.1] The section headings "3.2. Results of PSDNet" and "3.3. Results of DPDNet" should be renumbered to reflect the current section (e.g., 4.2 and 4.3), since Section 4 contains the datasets and results.
  2. [Tables 1–8] Table numbering is duplicated: there are two tables labeled "Table 3" (DPD block versus DPDNet body architecture) and two labeled "Table 4" (ImageNet architecture versus PSDNet results). Renumber all tables sequentially.
  3. [Section 3.2.1] "ResNet50 [20]" should cite the original ResNet architecture [8] rather than the identity-mapping paper [20], since the architecture used is the standard ResNet50.
  4. [Abstract and Conclusion] Several typographical errors should be corrected: "concolutional" and "netwok" in the abstract, "PDP inversed bottleneck block" in the conclusion, and "NerurIPS" in references [4] and [23] should be "NeurIPS".
  5. [Figure 2] The subfigure labels (a)–(d) are not explicitly referenced in the caption text beyond the order of the displayed blocks; please add a sentence identifying each block, and clarify whether linear bottlenecks (no ReLU after the PWC layer) are used in the DPD block as they are in MobileNetV2.
  6. [Table 5 and 6] The note "The number of parameters is the model for CIFAR-10 and CINIC-10 datasets" is ambiguous; for CIFAR-100 the FC layer has more output neurons, so the parameter count differs. Please report all three parameter counts explicitly or state the convolutional parameter count separately.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the efficiency comparison is closed-form arithmetic and the accuracy claims are direct empirical tests.

full rationale

The central efficiency claim that DWC channel expansion is cheaper than PWC expansion is derived in Section 3.1 from explicit parameter and FLOP counts; Eqs. (1) and (2) give the same closed-form ratio C/k^2 for both parameters and computation, with no fitted constants and no quantity being redefined as its own output. The accuracy claims for PSDNet and DPDNet are empirical comparisons against ResNet and MobileNetV2 on CIFAR-10, CIFAR-100, and CINIC-10, and the m and alpha sweeps are trade-off explorations rather than fits that force the headline result. There is a minor self-citation in related work (reference [15] includes author M. Zhang), but it is a survey citation used only to point to existing efficient-CNN literature, not a load-bearing premise. The ImageNet section contains placeholder numbers, but that is incomplete evaluation rather than circular reasoning. Therefore no step in the derivation reduces to its own inputs.

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

The paper's central argument rests on an arithmetic parameter-count comparison (Eq 1-2) plus empirical accuracy tables. No fitted constants are used. The main assumptions are that C >> k^2 in practical CNNs and that FLOPs and parameter counts proxy real efficiency; the paper also assumes DWC can substitute for PWC without a significant accuracy penalty.

assumptions (3)
  • domain assumption The number of input channels C is much greater than k squared, so the parameter-count ratio C/k^2 is much greater than 1 for practical CNNs.
    Invoked after Eq (1) and (2) to argue DWC is more efficient than PWC for channel expansion; if C is less than k^2, PWC would be cheaper.
  • domain assumption FLOPs and parameter counts are valid proxies for practical computational cost.
    The paper draws conclusions about computational cost from count arithmetic without runtime measurements on actual hardware.
  • domain assumption Replacing the channel-expansion pointwise convolution with a depthwise convolution does not harm the network's ability to combine channel information enough to lose accuracy.
    The PSD and DPD block designs assume DWC can substitute for PWC while preserving or improving accuracy; this is tested empirically but not guaranteed by the arithmetic.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PSDNet and DPDNet: Efficient channel expansion, Depthwise-Pointwise-Depthwise Inverted Bottleneck Block." pith.science (2026). https://pith.science/paper/JCDKA3RK

@misc{pith2026190901026,
  author       = {Pith},
  title        = {Pith review of: PSDNet and DPDNet: Efficient channel expansion, Depthwise-Pointwise-Depthwise Inverted Bottleneck Block},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JCDKA3RK}},
  note         = {Machine review of arXiv:1909.01026}
}
read the original abstract

In many real-time applications, the deployment of deep neural networks is constrained by high computational cost and efficient lightweight neural networks are widely concerned. In this paper, we propose that depthwise convolution (DWC) is used to expand the number of channels in a bottleneck block, which is more efficient than 1 x 1 convolution. The proposed Pointwise-Standard-Depthwise network (PSDNet) based on channel expansion with DWC has fewer number of parameters, less computational cost and higher accuracy than corresponding ResNet on CIFAR datasets. To design more efficient lightweight concolutional neural netwok, Depthwise-Pointwise-Depthwise inverted bottleneck block (DPD block) is proposed and DPDNet is designed by stacking DPD block. Meanwhile, the number of parameters of DPDNet is only about 60% of that of MobileNetV2 for networks with the same number of layers, but can achieve approximated accuracy. Additionally, two hyperparameters of DPDNet can make the trade-off between accuracy and computational cost, which makes DPDNet suitable for diverse tasks. Furthermore, we find the networks with more DWC layers outperform the networks with more 1x1 convolution layers, which indicates that extracting spatial information is more important than combining channel information.

Figures

Figures reproduced from arXiv: 1909.01026 by the authors.

Figure 1
Figure 1. (a). ResNet [20] uses 1×1 convolution to increase and decrease the number of channels so that the number of parameters does not increase explosively. When the DWC is adopted, the channel-multiplier (m) is usually set as 1 in most CNNs [18, 22, 34], which means that the number of input channels is the same as the number of output channels [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Comparison of bottleneck blocks for different architectures. The DPDNet block has two DWC layers and one PWC layer, and the first DWC layer implements channel expansion [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 37 canonical work pages

  1. [1]

    F. Nian, T. Li, Y . Wang, M. Xu, and J. Wu, Pornographic image detection utilizing deep convolutional neural networks, Neurocomputing 210 (2016) 283-293

    work page 2016

  2. [2]

    Zhang, W

    T. Zhang, W. Zheng, Z. Cui, Y . Zong, J. Yan, and K. Yan, A deep neural network -driven feature learning method for multi -view facial expression recognition, IEEE Trans. Multimedia 18 (2016) 2528-2536

    work page 2016

  3. [3]

    W. Liu, Z. Wang, X. Liu, N. Zeng, Y . Li u, and F. E. Alsaadi, A survey of deep neural network architectures and their applications, Neurocomputing 234 (2017) 11-26

    work page 2017

  4. [4]

    Krizhevsky, I

    A. Krizhevsky, I. Sutskever, and G. E. Hinton, Imagenet classification with deep convolutional neural networks, in: Advances in neural information processing systems (NerurIPS), 2012, 1097-1105

    work page 2012

  5. [5]

    Russakovsky, J

    O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A. Khosla, M. Bernstein, A. C. Berg, and F. -F. Li, Imagenet large scale visual recognition challen ge, Int. J. Comput. Vision 115 (2015) 211-252

    work page 2015

  6. [6]

    Simonyan and A

    K. Simonyan and A. Zisserman, Very deep convolutional networks for large -scale image recognition, in: International Conference on Learning Representations (ICLR), 2014

    work page 2014

  7. [7]

    Szegedy, W

    C. Szegedy, W. Liu, Y . Jia, P. Sermanet, S. Reed, D. Anguelov, D. Erhan, V . Vanhoucke, and A. Rabinovich, Going deeper with convolutions, in: Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), 2015, 1-9

    work page 2015

  8. [8]

    K. He, X. Zhang, S. Ren, and J. Sun, Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), 2016, 770 -778

    work page 2016

Show all 48 references
  1. [9]

    Huang, Z

    G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Weinberger, Densely Connected Convolutional Networks, in: IEEE conference on computer vision and pattern recognition (CVPR), 2017, 3

    2017

  2. [10]

    Zagoruyko and N

    S. Zagoruyko and N. Komodakis, Wide residual networks, arXiv preprint arXiv: 1605.07146 (2016)

    2016 arXiv

  3. [11]

    S. Xie, R. Girshick, P. Dollá r, Z. Tu, and K. He, Aggregated residual transformations for deep neural networks, in: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 5987-5995

    2017

  4. [12]

    Jiang, Y

    X. Jiang, Y . Pang, X. Li, J. Pan, and Y . Xie, Deep neural networks with elastic rectified linear units for object recognition, Neurocomputing 275 (2018) 1132-1139

    2018

  5. [13]

    G. Zhao, G. Liu, L. Fang, B. Tu, and P. Ghamisi, Multiple convolutional layers fusion framework for hyperspectral image classification, Neurocomputing 339 (2019) 149 -160

    2019

  6. [14]

    Brahimi, N

    S. Brahimi, N. B. Aoun, and C. B. Amar, Boosted Convolutional Neural Network for object recognition at large scale, Neurocomputing 330 (2019) 337-354

    2019

  7. [15]

    Zhang, M

    Q. Zhang, M. Zhang, T. Chen, Z. Sun, Y . Ma, and B. Yu, Recent advances in convolutional neural network acceleration, Neurocomputing 323 (2019) 37-51

    2019

  8. [16]

    Z. Liu, M. Sun, T. Zhou, G. Huang, and T. Darrell, Rethinking the Value of Network Pruning, arXiv preprint arXiv:1810.05270 (2018)

    2018 arXiv

  9. [17]

    S. Han, H. Mao, and W. J. Dally, Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding, in: International Conference on Learning Representations (ICLR), 2016

    2016

  10. [18]

    A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang, T. Weyand, M. Andreetto, and H. Adam, Mobilenets: Efficient convolutional neural n etworks for mobile vision applications, arXiv preprint arXiv:1704.04861 (2017). 14

    2017 arXiv

  11. [19]

    Megvii, ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices, arXiv: 1707.01083 (20 17)

    Zhang Xiangyu, Zhou Xinyu, Lin Mengxiao, Sun Jian, and I. Megvii, ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices, arXiv: 1707.01083 (20 17)

    arXiv

  12. [20]

    K. He, X. Zhang, S. Ren, and J. Sun, Identity mappings in deep residual networks, in: European conference on computer vision (ECCV), 2016, 630-645

    2016

  13. [21]

    F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, Squeezenet: Alexnet-level accuracy with 50x fewer parameters and< 0.5 mb model size, arXiv preprint arXiv: 1602.07360 (2016)

    2016 arXiv

  14. [22]

    Sandler, A

    M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, MobileNetV2: Inverted Residuals and Linear Bottlenecks, in: Proceedings of th e IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, 4510-4520

    2018

  15. [23]

    S. Han, J. Pool, J. Tran, and W. J. Dally, Learning both weights and connections for efficient neural network, in: Advances in neural information processing systems (NerurIPS), 2015, 1135-1143

    2015

  16. [24]

    H. Li, A. Kadav, I. Durdanovic, H. Samet, and H. P. Graf, Pruning filters for efficient convnets, in: International Conference on Learning Representations (ICLR), 2017

    2017

  17. [25]

    Y . He, X. Zhang, and J. Sun, Channel pruning for accelerating very deep neural networks, in: International Conference on Computer Vision (ICCV), 2017

    2017

  18. [26]

    Courbariaux, Y

    M. Courbariaux, Y . Bengio, and J.-P. David, Binaryconnect: Training deep neural networks with binary weights during propagations, in: Advances in ne ural information processing systems (NerurIPS), 2015, 3123-3131

    2015

  19. [27]

    Courbariaux, I

    M. Courbariaux, I. Hubara, D. Soudry, R. El -Yaniv, and Y . Bengio, Binarized neural networks: Training deep neural networks with weights and activations constrained to+ 1 or -1, arXiv preprint arXiv: 1602.02830 (2016)

    2016 arXiv

  20. [28]

    Rastegari, V

    M. Rastegari, V . Ordonez, J. Redmon, and A. Farhadi, Xnor -net: Imagenet classification using binary convolutional neural networks, in: European Conference on Computer Vision (ECCV), 2016, 525-542

    2016

  21. [29]

    F. Li, B. Zhang, and B. Liu, Ternary Weight Networks, arXiv: 1605.04711 (2016)

    2016 arXiv

  22. [30]

    Alemdar, V

    H. Alemdar, V . Leroy, A. Prost-Boucle, and F. Pé trot, Ternary neural networks for resource-efficient AI applications, in: 2017 international joint conference on neural networks (IJCNN), 201 7, 2547-2554

    2017

  23. [31]

    Zhang, G.-J

    T. Zhang, G.-J. Qi, B. Xiao, and J. Wang, Interleaved group convolutions, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 4373-4382

    2017

  24. [32]

    Ioffe and C

    S. Ioffe and C. Szegedy, Batch normalization: Accel erating deep network training by reducing internal covariate shift, arXiv preprint arXiv: 1502.03167 (2015)

    2015 arXiv

  25. [33]

    Huang, S

    G. Huang, S. Liu, L. van der Maaten, and K. Q. Weinberger, CondenseNet: An Efficient DenseNet using Learned Group Convolutions, arXiv preprint arXiv: 1711.09224 (2017)

    2017 arXiv

  26. [34]

    Chollet, Xception: Deep learning with depthwise separable convolutions, in: IEEE conference on computer vision and pattern recognition (CVPR), 2017

    F. Chollet, Xception: Deep learning with depthwise separable convolutions, in: IEEE conference on computer vision and pattern recognition (CVPR), 2017

    2017

  27. [35]

    J. Jin, A. Dundar, and E. J. a. p. a. Culurciello, Flattened convolutional neural networks for feedforward acceleration, in: International Conference on Learning Representations (ICLR), 2014

    2014

  28. [36]

    Szegedy, S

    C. Szegedy, S. Ioffe, V . Vanhoucke, and A. A. Alemi, Inception-v4, inception-resnet and the impact of residual connectio ns on learning, in: Association for the Advance of Artificial Intelligence (AAAI), 2017, 12. 15

    2017

  29. [37]

    Szegedy, V

    C. Szegedy, V . Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, Rethinking the inception architecture for computer vision, in: Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), 2016, 2818-2826

    2016

  30. [38]

    Ou and Y

    J. Ou and Y . Li, Vector-kernel Convolutional Neural Networks, Neurocomputing (2018)

    2018

  31. [39]

    G. E. Dahl, T. N. Sainath, and G. E. Hinton, Improving deep neural networks for LVCSR using rectified linear units and dropout, in: 2013 IEEE International Conference on Speech and Signal Processing (ICASSP), 2013, 8609-8613

    2013

  32. [40]

    Krizhevsky and G

    A. Krizhevsky and G. Hinton, Learning multiple layers of features from tiny images, Technical report (2009)

    2009

  33. [41]

    Torralba, R

    A. Torralba, R. Fergus, and W. T. Freeman, 80 million tiny images: A large data set for nonparametric object and scene recognition, IEEE Trans. pattern Anal. Mach. Intell. 30 (2008) 1958-1970

    2008

  34. [42]

    Huang, Y

    G. Huang, Y . Sun, Z. Liu, D. Sedra, and K. Q. Weinberger, Deep networks with stochastic depth, in: European Conference on Computer Vision (ECCV), 2016, 646-661

    2016

  35. [43]

    Larsson, M

    G. Larsson, M. Maire, and G. J. a. p. a. Shakhnarovich, Fractalnet: Ultra -deep neural networks without residuals, in: International Conference on Learning Representations (ICLR), 2016

    2016

  36. [44]

    K. Sun, M. Li, D. Liu, and J. Wang, IGCV3: Interleaved Low -Rank Group Convolutions for Efficient Deep Neural Networks, arXiv preprint arXiv: 1806.00178 (2018)

    2018 arXiv

  37. [45]

    G. Xie, J. Wang, T. Zhang, J. Lai, R. Hong, and G. -J. Qi, IGCV 2 : Interleaved Structured Sparse Convolutional Neural Networks, arXiv preprint arXiv: 1804.06202 (2018)

    2018 arXiv

  38. [46]

    M. Lin, Q. Chen, and S. Yan, Network in network, arXiv: 1213.4400 (2013)

    2013

  39. [47]

    L. N. Darlow, E. J. Crowley, A. Antoniou, and A. J. Sto rkey, CINIC -10 is not ImageNet or CIFAR-10, arXiv preprint arXiv: 1810.03505 (2018)

    2018 arXiv

  40. [48]

    Abadi, P

    M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, and M. Isard, Tensorflow: a system for large -scale machine learning, in: P roceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI), 2016, 265 -283

    2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.