Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

Electrostatic Force Regularization for Neural Structured Pruning

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that adding an electrostatic-force penalty to the training loss drives some convolutional filters' weights to zero and preserves others, so the network can be pruned at arbitrary ratios without retraining or architecture…

desk verdict A Coulomb-style pruning story that fails on its own math: the derivation is invalid and the update rule is just weighted L2 shrinkage, though the experimental setup is more substantial than the theory. read the letter →

arxiv 2411.11079 v1 pith:XA2R6W6O submitted 2024-11-17 cs.CV

Abdesselam Ferdi , Abdelmalik Taleb-Ahmed , Amir Nakib , Youcef Ferdi This is my paper · ORCID
classification cs.CV
keywords structuredpruningelectrostaticforceDCNNsfilterregularizationmodelcompressionnofine-tuning
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 proposes a structured-pruning method that treats each convolutional filter as a charged object. A source filter with the largest $\ell^1$-norm creates an electrostatic field: filters with the same charge are pushed toward zero weights, while filters with the opposite charge are pulled toward non-zero weights. After training with this penalty, the zero-weight filters can be deleted at any pruning ratio, with no architectural changes and no lengthy fine-tuning. The authors report pruned accuracies competitive with existing structured-pruning methods on ResNet and VGG models over MNIST, CIFAR, and ImageNet.

What carries the argument

The central object is the electrostatic force penalty $F_{e,n,l} = k_e |q_{1,l}| |q_{n,l}| / r_{n,l}^2$, with $q_{n,l} = \mathrm{sgn}(n,l) \|W_{n,l}\|_1$ and $r_{n,l} = |q_{1,l} - q_{n,l}|$. The source filter $q_1$ is the layer's largest-$\ell^1$-norm filter. The paper's claimed mechanism is the gradient update $w \leftarrow w - \epsilon(\nabla_w J + \alpha_e k_e |q_1| r^{-2} w)$, which supposedly drives same-sign filters to zero and opposite-sign filters to larger magnitudes. This penalty is what couples weight optimization with filter-importance ranking.

What would settle it

Train one convolutional layer on MNIST with the electrostatic regularizer and plot the fraction of repelled-filter $\ell^1$-norms that reach machine-precision zero over epochs. Independently, differentiate $F_{e,n,l}$ with respect to a weight and compare the result with the update rule in the paper: if the analytic gradient is not proportional to the weight itself (or if repelled norms plateau far from zero), the claimed mechanism fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that a Coulomb-style penalty added to the training objective reorganizes a convolutional layer's filters into two populations: filters whose charge matches the source filter (the filter with the largest $\ell^1$-norm) feel a repulsive force and their weights shrink toward zero, while filters of opposite charge feel an attractive force and keep non-zero weights. Concretely, the regularized loss is $\tilde{J} = \sum J + \alpha_e F_e$ with $F_{e,n,l} = k_e |q_{1,l}| |q_{n,l}| / r_{n,l}^2$, where charges are the product of filter sign and $\ell^1$-norm and $r$ is the absolute difference of charges. The authors claim that after training with this objective, deleting the repelled filters by $\ell^1$-norm ranking yields a pruned network with minimal accuracy loss, without architecture changes and without an extensive fine-tuning stage. They present the method as the first application of electrostatic charge concepts to DCNN training for pruning, and report competitive results on MNIST, CIFAR, and ImageNet across ResNet and VGG architectures.

Load-bearing premise

The pruning mechanism rests on the claim that minimizing the penalty term pushes repelled filters' weights to zero and attracted filters' weights to larger magnitudes; if the effective gradient does not have that direction, the clean separation of disposable and informative filters does not follow.

Editorial extensions

If this is right

  • A single training run yields a model that can be pruned at multiple ratios without retraining, because the regularization already separates disposable from informative filters.
  • Pruning removes whole filters rather than individual weights, so the gains translate directly into fewer FLOPs and less memory with no architectural changes.
  • The method works from both random and pretrained initialization, with pretrained weights giving higher pruned accuracy on the reported benchmarks.
  • On ImageNet, the reported $1.34\times$ speedup for ResNet-34 keeps the top-1 drop at $0.19\%$, a level the authors say beats or matches several baselines at larger speedups.

Reading between the lines

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

  • A natural extension is to apply the same charge-and-distance penalty to other structured units—transformer attention heads, MLP neurons, or convolutional blocks—wherever a per-unit $\ell^1$-norm can serve as the charge magnitude.
  • The claimed pruning-ratio flexibility suggests a workflow in which one electrostatic-trained model is deployed at different speed/accuracy trade-offs without retraining; this could be tested by training once and evaluating a ladder of pruning ratios on a held-out set.
  • Because the force magnitude depends on the difference of filter charges, layers with many filters of similar magnitude may require per-layer tuning of $\alpha_e$; an adaptive scheme that sets $\alpha_e$ from the spread of $\ell^1$-norms is a possible improvement the paper does not explore.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes a structured pruning method for deep convolutional neural networks in which an electrostatic-force-inspired regularizer is added to the training loss. The charge of a filter is defined as the sign of the sum of its weights multiplied by the L1 norm of the weights, the source filter is the filter with the largest L1 norm, and the force between filters is defined by Fe = ke |q1||q|/r^2, with r the absolute difference of the two scalar charges. The authors claim that minimizing this regularizer causes filters with like charges to shrink toward zero while filters with opposite charges grow toward non-zero values, allowing filters to be pruned after training without architectural changes or retraining. Experiments are reported on MNIST, CIFAR-10, CIFAR-100, and ImageNet with ResNet and VGG architectures, comparing pruned accuracy and speedup against several prior pruning methods.

Significance. If the proposed mechanism worked as described, the method would be practically attractive: it promises simple structured pruning at arbitrary ratios without fine-tuning and with no architectural modification. The paper also provides an interesting analogy between Coulomb's law and filter importance. However, the central mathematical derivation is incorrect: the gradient of the proposed regularizer is not what the paper claims, and the update rule in Eq. 11 does not realize the advertised attraction/repulsion behavior. Because the pruning mechanism is unsupported and no code or targeted ablations are provided to verify the claimed force dynamics, the empirical results cannot establish the central claim. The paper does include comparisons to several published pruning methods and reports training overhead, which are useful, but these do not compensate for the unsupported mechanism.

major comments (4)
  1. [Section 3.2, Eqs. (6)-(11)] The derivation from Eq. 6 to Eq. 10 is not a valid gradient computation. From the definition Fe = ke |q1||q|/r^2, with |q| = ||W||_1 and r = |q1 - q|, the partial derivative of Fe with respect to a weight w_i is not proportional to w_i. The derivative of the L1 norm is sign(w_i) (where defined), and the distance r depends on q, so its derivative contributes additional terms through dq/dw_i. Therefore Eq. 10, and consequently the update rule in Eq. 11, do not follow from the stated objective.
  2. [Section 3.2, Eq. 11] Even if Eq. 10 were taken as the implemented update rather than as a derivative, the force term alpha_e ke |q1|/r^2 (w) is a shrinkage term proportional to w: it pushes all weights toward zero, with the magnitude of the push depending on the distance r. It contains no signed direction that would pull opposite-sign filters toward non-zero weights. The 'attractive' case is at most a weaker shrinkage when r is large, not a force toward non-zero values. Thus the paper's central claim that same-sign filters are repelled to zero while opposite-sign filters are attracted to non-zero values is not supported by the equations as written.
  3. [Section 6.4 and Table 2] The hyperparameter alpha_e is selected per model and dataset after observing the pruning curves. Section 6.4 states that the 'appropriate' values were determined from Figure 2, and Table 2 uses different alpha_e values for ResNet-56, VGG-19, and ResNet-34. Because alpha_e controls the strength of the alleged force and is tuned to produce good pruning results, the reported comparisons partly demonstrate hyperparameter fitting rather than a validated mechanism. Without an ablation that isolates the claimed attraction/repulsion behavior or code to reproduce the force dynamics, the experiments cannot confirm the mechanism.
  4. [Section 6.3.1, Table 4] The claim that the method yields 'superior top-1 accuracy after pruning compared to all existing methods' is overstated because the comparison groups methods with different speedups and different baseline accuracies. For example, at 2.17x speedup the electrostatic force (p) model reaches 93.88% pruned accuracy, but Torque at 2.15x reaches 93.76% and ABC Pruner at 2.18x reaches 93.23%; at 2.62x the electrostatic force (r) model drops to 93.04%, which is worse than several methods at similar speedups. A direct comparison under matched speedup and baseline is needed to support the comparative claim.
minor comments (5)
  1. [Section 1] The sentence beginning 'In the rest of this paper. In Section 2...' contains a period after 'paper' and should be rewritten as a single grammatical sentence.
  2. [Section 6.1] The value '74,49%' uses a comma as a decimal separator, which is inconsistent with the decimal-point notation used throughout the rest of the manuscript.
  3. [Algorithm 1] The line 'if n = source_f ilter_index or (nl) = 0then' contains a formatting artifact and should read 'if n = source_filter_index or sign(q_n,l) = 0 then'.
  4. [Figure 2 caption and Section 6.4] The Figure 2 caption mentions 'four distinct electrostatic force rates', while Section 6.4 reports experiments with three values of alpha_e (10^-11, 10^-12, and 10^-14). The manuscript should reconcile this discrepancy.
  5. [Author list] The author name 'Y oucef Ferdi' contains a stray space and should be corrected to 'Youcef Ferdi'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the electrostatic-force pruning results are evaluated empirically against external benchmarks; the flawed Eq. 6-to-Eq. 10 gradient step is a mathematical-correctness issue, not a circularity.

full rationale

The paper's derivation chain is not circular: the 'charge' is defined from filter weights (Eqs. 2-4), and the force from that charge (Eq. 6), but the claimed pruning behavior is then tested empirically on MNIST/CIFAR/ImageNet (Tables 4-6) rather than being read off from the definition alone. There are no load-bearing self-citations; the reference list contains no works by the authors. Alpha_e is a tuned hyperparameter (Fig. 2, Table 2), but the evaluated quantities are measured pruned accuracies, not quantities constructed from the fitted values. The serious weakness is that Eq. 10 does not follow from Eq. 6: for |q| = ||W||_1, d|q|/dw_i = sign(w_i), and r in Eq. 5 depends on W, so the gradient is not proportional to w; and Eq. 11's 'attractive' case is at most a weaker shrinkage, not a force toward non-zero weights. That is an unsupported derivation / correctness risk, not a circular equivalence between input and output, so it does not raise the circularity score.

Assumptions & free parameters 2 free parameters · 4 assumptions · 2 invented entities

The paper's central mechanism rests on two constructed quantities, the filter charge and the inter-filter force, neither of which has independent empirical support. The force rate and pruning ratios are tuned per experiment. The main mathematical premise, that the gradient of F_e points along the weight vector, is asserted in Eqs. 10 and 11 and is false under the definition in Eq. 6.

free parameters (2)
  • alpha_e (electrostatic force rate) = 10^-16 for ResNet-56/CIFAR-10, 10^-11 for VGG-19/CIFAR-100, 10^-17 for ResNet-34/ImageNet
    Chosen per model and dataset by ablation; controls the strength of the penalty and directly determines the pruning behavior.
  • per-layer pruning ratios = [0,0.52,0.52,0.52,0], [0,0.6,0.6,0.6,0], [0,0.62,0.63,0.62,0], [0:0,1-15:0.65], [0:0,1-15:0.70], [0,0.5,0.5,0.5,0,0]
    Set manually per experiment to hit target speedups; accuracy results are reported for these chosen ratios only.
assumptions (4)
  • domain assumption Filter importance can be ranked by the L1 norm of its weights, and the largest-norm filter can serve as the source filter.
    The pruning decisions and the force field both depend on this heuristic, which is asserted rather than justified.
  • ad hoc to paper The sign of the sum of the filter weights defines the electrostatic charge polarity of that filter.
    Introduced solely to create the repulsion and attraction categorization; no independent evidence links this quantity to filter importance.
  • ad hoc to paper The distance between filters can be represented as the absolute difference of their scalar charges.
    This scalar distance definition is chosen to make the force formula simple and is not justified by the structure of convolutional filters.
  • ad hoc to paper Minimizing F_e through gradient descent produces the desired weight distribution.
    The paper assumes in Eqs. 10 and 11 that the gradient of F_e is proportional to the weight vector, but this is false for the defined F_e, so the premise is unsupported.
invented entities (2)
  • Filter charge q_n,l
    purpose: Assign a scalar sign and magnitude to each convolution filter so a Coulomb-like force can be computed.
    No physical or empirical basis; it is a constructed scalar from the sign of the weight sum and the L1 norm.
  • Electrostatic force F_e between filters
    purpose: Added to the training loss to push filters toward zero or preserve them.
    A mathematical construct whose gradient as written does not match its definition, with no external measurements confirming its effect.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Electrostatic Force Regularization for Neural Structured Pruning." pith.science (2026). https://pith.science/paper/XA2R6W6O

@misc{pith2026241111079,
  author       = {Pith},
  title        = {Pith review of: Electrostatic Force Regularization for Neural Structured Pruning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XA2R6W6O}},
  note         = {Machine review of arXiv:2411.11079}
}
read the original abstract

The demand for deploying deep convolutional neural networks (DCNNs) on resource-constrained devices for real-time applications remains substantial. However, existing state-of-the-art structured pruning methods often involve intricate implementations, require modifications to the original network architectures, and necessitate an extensive fine-tuning phase. To overcome these challenges, we propose a novel method that, for the first time, incorporates the concepts of charge and electrostatic force from physics into the training process of DCNNs. The magnitude of this force is directly proportional to the product of the charges of the convolution filter and the source filter, and inversely proportional to the square of the distance between them. We applied this electrostatic-like force to the convolution filters, either attracting filters with opposite charges toward non-zero weights or repelling filters with like charges toward zero weights. Consequently, filters subject to repulsive forces have their weights reduced to zero, enabling their removal, while the attractive forces preserve filters with significant weights that retain information. Unlike conventional methods, our approach is straightforward to implement, does not require any architectural modifications, and simultaneously optimizes weights and ranks filter importance, all without the need for extensive fine-tuning. We validated the efficacy of our method on modern DCNN architectures using the MNIST, CIFAR, and ImageNet datasets, achieving competitive performance compared to existing structured pruning approaches.

Figures

Figures reproduced from arXiv: 2411.11079 by the authors.

Figure 1
Figure 1. An illustration of electrostatic force-based training. A convolutional layer with ten filters is used as an example. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Top-1 accuracy of pruned ResNet-56 and VGG-19 models, initialized with pretrained weights and trained [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Normalized filter L1-norm of layers 1 and 2 for the ResNet-56/CIFAR-10 model trained with and without electrostatic force, and with L1-norm. 6.3 Comparison With State-of-the-arts The subsequent analysis will focus on the model accuracy after pruning and the acceleration achieved, which is quantified by the reduction in FLOPs. This is consistent with our objective of proposing a SP method that can be used to create a… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Lightweight G-YOLOv11: Advancing Efficient Fracture Detection in Pediatric Wrist X-rays

    eess.IV 2024-12 conditional novelty 3.0 of 10

    G-YOLOv11l, a ghost-convolution and halved-filter YOLOv11 variant, reaches mAP@0.5 of 0.535 at 2.4 ms per image, 68.7% smaller than YOLOv11l but 13.6% less accurate.

  2. Deep Convolutional Neural Networks Structured Pruning via Gravity Regularization

    cs.CV 2024-11 reject novelty 2.0 of 10

    A gravity-inspired regularization method for structured pruning is proposed, but its no-retraining results show large accuracy drops and its fine-tuned comparisons use lower baselines than competing methods.

Reference graph

Works this paper leans on

38 extracted references · 25 canonical work pages · cited by 2 Pith papers

  1. [1]

    Restructuring the teacher and student in self-distillation

    Yujie Zheng, Chong Wang, Chenchen Tao, Sunqi Lin, Jiangbo Qian, and Jiafei Wu. Restructuring the teacher and student in self-distillation. IEEE Transactions on Image Processing, 2024

    work page 2024

  2. [2]

    Low-rank approximation for sparse attention in multi-modal llms

    Lin Song, Yukang Chen, Shuai Yang, Xiaohan Ding, Yixiao Ge, Ying-Cong Chen, and Ying Shan. Low-rank approximation for sparse attention in multi-modal llms. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13763–13773, 2024

    work page 2024

  3. [3]

    Quantization via distillation and contrastive learning

    Zehua Pei, Xufeng Yao, Wenqian Zhao, and Bei Yu. Quantization via distillation and contrastive learning. IEEE Transactions on Neural Networks and Learning Systems, 2023

    2023

  4. [4]

    Discrimination-aware network pruning for deep model compression

    Jing Liu, Bohan Zhuang, Zhuangwei Zhuang, Yong Guo, Junzhou Huang, Jinhui Zhu, and Mingkui Tan. Discrimination-aware network pruning for deep model compression. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(8):4035–4051, 2021

    work page 2021

  5. [5]

    Complexity-driven model compression for resource- constrained deep learning on edge

    Muhammad Zawish, Steven Davy, and Lizy Abraham. Complexity-driven model compression for resource- constrained deep learning on edge. IEEE Transactions on Artificial Intelligence, 5(8):3886–3901, 2024

    work page 2024

  6. [6]

    Ganji, Ivan Lazarevich, and Sudhakar Sah

    Matteo Grimaldi, Darshan C. Ganji, Ivan Lazarevich, and Sudhakar Sah. Accelerating deep neural networks via semi-structured activation sparsity. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, pages 1179–1188, October 2023

    work page 2023

  7. [7]

    Advancing model pruning via bi-level optimization

    Yihua Zhang, Yuguang Yao, Parikshit Ram, Pu Zhao, Tianlong Chen, Mingyi Hong, Yanzhi Wang, and Sijia Liu. Advancing model pruning via bi-level optimization. Advances in Neural Information Processing Systems, 35:18309–18326, 2022

    work page 2022

  8. [8]

    Prior gradient mask guided pruning-aware fine-tuning

    Linhang Cai, Zhulin An, Chuanguang Yang, Yangchun Yan, and Yongjun Xu. Prior gradient mask guided pruning-aware fine-tuning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pages 140–148, 2022

    work page 2022

Show all 38 references
  1. [9]

    The lottery ticket hypothesis: Finding sparse, trainable neural networks

    Jonathan Frankle and Michael Carbin. The lottery ticket hypothesis: Finding sparse, trainable neural networks. In International Conference on Learning Representations, 2019

    2019

  2. [10]

    Rethinking the value of network pruning

    Zhuang Liu, Mingjie Sun, Tinghui Zhou, Gao Huang, and Trevor Darrell. Rethinking the value of network pruning. arXiv preprint arXiv:1810.05270, 2018. 11

    2018 arXiv

  3. [11]

    Snip: Single-shot network pruning based on connection sensitivity

    Namhoon Lee, Thalaiyasingam Ajanthan, and Philip Torr. Snip: Single-shot network pruning based on connection sensitivity. In International Conference on Learning Representations, 2018

    2018

  4. [12]

    Progressive skeletonization: Trimming more fat from a network at initialization

    Pau de Jorge, Amartya Sanyal, Harkirat Behl, Philip Torr, Grégory Rogez, and Puneet K Dokania. Progressive skeletonization: Trimming more fat from a network at initialization. In International Conference on Learning Representations, 2021

    2021

  5. [13]

    Picking winning tickets before training by preserving gradient flow

    Chaoqi Wang, Guodong Zhang, and Roger Grosse. Picking winning tickets before training by preserving gradient flow. arXiv preprint arXiv:2002.07376, 2020

    2002 arXiv

  6. [14]

    Linear mode connectivity and the lottery ticket hypothesis

    Jonathan Frankle, Gintare Karolina Dziugaite, Daniel Roy, and Michael Carbin. Linear mode connectivity and the lottery ticket hypothesis. In International Conference on Machine Learning, pages 3259–3269. PMLR, 2020

    2020

  7. [15]

    Neural pruning via growing regularization

    Huan Wang, Can Qin, Yulun Zhang, and Yun Fu. Neural pruning via growing regularization. In International Conference on Learning Representations (ICLR), 2021

    2021

  8. [16]

    Pruning parameterization with bi-level optimization for efficient semantic segmentation on the edge

    Changdi Yang, Pu Zhao, Yanyu Li, Wei Niu, Jiexiong Guan, Hao Tang, Minghai Qin, Bin Ren, Xue Lin, and Yanzhi Wang. Pruning parameterization with bi-level optimization for efficient semantic segmentation on the edge. In Proceedings of the IEEE/CVF Conference on Computer Vision ...

    2023

  9. [17]

    Gradual channel pruning while training using feature relevance scores for convolutional neural networks

    Sai Aparna Aketi, Sourjya Roy, Anand Raghunathan, and Kaushik Roy. Gradual channel pruning while training using feature relevance scores for convolutional neural networks. IEEE Access, 8:171924–171932, 2020

    2020

  10. [18]

    Learning filter pruning criteria for deep convolutional neural networks acceleration

    Yang He, Yuhang Ding, Ping Liu, Linchao Zhu, Hanwang Zhang, and Yi Yang. Learning filter pruning criteria for deep convolutional neural networks acceleration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2009–2018, 2020

    2009

  11. [19]

    Structured compression of deep neural networks with debiased elastic group lasso

    Oyebade Oyedotun, Djamila Aouada, and Bjorn Ottersten. Structured compression of deep neural networks with debiased elastic group lasso. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 2277–2286, 2020

    2020

  12. [20]

    Rigging the lottery: Making all tickets winners

    Utku Evci, Trevor Gale, Jacob Menick, Pablo Samuel Castro, and Erich Elsen. Rigging the lottery: Making all tickets winners. In International Conference on Machine Learning, pages 2943–2952. PMLR, 2020

    2020

  13. [21]

    Efficient joint optimization of layer-adaptive weight pruning in deep neural networks

    Kaixin Xu, Zhe Wang, Xue Geng, Min Wu, Xiaoli Li, and Weisi Lin. Efficient joint optimization of layer-adaptive weight pruning in deep neural networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 17447–17457, 2023

    2023

  14. [22]

    Torque based structured pruning for deep neural network

    Arshita Gupta, Tien Bau, Joonsoo Kim, Zhe Zhu, Sumit Jha, and Hrishikesh Garud. Torque based structured pruning for deep neural network. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 2711–2720, 2024

    2024

  15. [23]

    Orthcaps: An orthogonal capsnet with sparse attention routing and pruning

    Xinyu Geng, Jiaming Wang, Jiawei Gong, Yuerong Xue, Jun Xu, Fanglin Chen, and Xiaolin Huang. Orthcaps: An orthogonal capsnet with sparse attention routing and pruning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6037–6046, 2024

    2024

  16. [24]

    Finding lottery tickets in vision models via data- driven spectral foresight pruning

    Leonardo Iurada, Marco Ciccone, and Tatiana Tommasi. Finding lottery tickets in vision models via data- driven spectral foresight pruning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16142–16151, 2024

    2024

  17. [25]

    Bilevelpruning: Unified dynamic and static channel pruning for convolutional neural networks

    Shangqian Gao, Yanfu Zhang, Feihu Huang, and Heng Huang. Bilevelpruning: Unified dynamic and static channel pruning for convolutional neural networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16090–16100, 2024

    2024

  18. [26]

    Unipts: A unified framework for proficient post-training sparsity

    Jingjing Xie, Yuxin Zhang, Mingbao Lin, Liujuan Cao, and Rongrong Ji. Unipts: A unified framework for proficient post-training sparsity. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5746–5755, 2024

    2024

  19. [27]

    Channel pruning for accelerating very deep neural networks

    Yihui He, Xiangyu Zhang, and Jian Sun. Channel pruning for accelerating very deep neural networks. In Proceedings of the IEEE international conference on computer vision, pages 1389–1397, 2017

    2017

  20. [28]

    Amc: Automl for model compression and acceleration on mobile devices

    Yihui He, Ji Lin, Zhijian Liu, Hanrui Wang, Li-Jia Li, and Song Han. Amc: Automl for model compression and acceleration on mobile devices. In Proceedings of the European conference on computer vision (ECCV), pages 784–800, 2018

    2018

  21. [29]

    Filter pruning via geometric median for deep convolutional neural networks acceleration

    Yang He, Ping Liu, Ziwei Wang, Zhilan Hu, and Yi Yang. Filter pruning via geometric median for deep convolutional neural networks acceleration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4340–4349, 2019

    2019

  22. [30]

    Soft filter pruning for accelerating deep convolutional neural networks

    Yang He, Guoliang Kang, Xuanyi Dong, Yanwei Fu, and Yi Yang. Soft filter pruning for accelerating deep convolutional neural networks. arXiv preprint arXiv:1808.06866, 2018. 12

    2018 arXiv

  23. [31]

    Whc: Weighted hybrid criterion for filter pruning on convolutional neural networks

    Shaowu Chen, Weize Sun, and Lei Huang. Whc: Weighted hybrid criterion for filter pruning on convolutional neural networks. In ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1–5. IEEE, 2023

    2023

  24. [32]

    Channel pruning via automatic structure search

    Mingbao Lin, Rongrong Ji, Yuxin Zhang, Baochang Zhang, Yongjian Wu, and Yonghong Tian. Channel pruning via automatic structure search. arXiv preprint arXiv:2001.08565, 2020

    2001 arXiv

  25. [33]

    Channel pruning via lookahead search guided reinforcement learning

    Zi Wang and Chengcheng Li. Channel pruning via lookahead search guided reinforcement learning. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 2029–2040, 2022

    2022

  26. [34]

    Centripetal sgd for pruning very deep convolu- tional networks with complicated structure

    Xiaohan Ding, Guiguang Ding, Yuchen Guo, and Jungong Han. Centripetal sgd for pruning very deep convolu- tional networks with complicated structure. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4943–4953, 2019

    2019

  27. [35]

    Auto-balanced filter pruning for efficient convolutional neural networks

    Xiaohan Ding, Guiguang Ding, Jungong Han, and Sheng Tang. Auto-balanced filter pruning for efficient convolutional neural networks. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 32, 2018

    2018

  28. [36]

    Eigendamage: Structured pruning in the kronecker-factored eigenbasis

    Chaoqi Wang, Roger Grosse, Sanja Fidler, and Guodong Zhang. Eigendamage: Structured pruning in the kronecker-factored eigenbasis. In International conference on machine learning, pages 6566–6575. PMLR, 2019

    2019

  29. [37]

    Importance estimation for neural network pruning

    Pavlo Molchanov, Arun Mallya, Stephen Tyree, Iuri Frosio, and Jan Kautz. Importance estimation for neural network pruning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11264–11272, 2019

    2019

  30. [38]

    Pruning filters for efficient convnets

    Hao Li, Asim Kadav, Igor Durdanovic, Hanan Samet, and Hans Peter Graf. Pruning filters for efficient convnets. arXiv preprint arXiv:1608.08710, 2016. 13

    2016 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.