|
|
|
|
| LEADER |
01000naa a22002652 4500 |
| 001 |
NLM344371719 |
| 003 |
DE-627 |
| 005 |
20231226022620.0 |
| 007 |
cr uuu---uuuuu |
| 008 |
231226s2023 xx |||||o 00| ||eng c |
| 024 |
7 |
|
|a 10.1109/TPAMI.2022.3195774
|2 doi
|
| 028 |
5 |
2 |
|a pubmed24n1147.xml
|
| 035 |
|
|
|a (DE-627)NLM344371719
|
| 035 |
|
|
|a (NLM)35917571
|
| 040 |
|
|
|a DE-627
|b ger
|c DE-627
|e rakwb
|
| 041 |
|
|
|a eng
|
| 100 |
1 |
|
|a Lin, Mingbao
|e verfasserin
|4 aut
|
| 245 |
1 |
0 |
|a 1xN Pattern for Pruning Convolutional Neural Networks
|
| 264 |
|
1 |
|c 2023
|
| 336 |
|
|
|a Text
|b txt
|2 rdacontent
|
| 337 |
|
|
|a ƒaComputermedien
|b c
|2 rdamedia
|
| 338 |
|
|
|a ƒa Online-Ressource
|b cr
|2 rdacarrier
|
| 500 |
|
|
|a Date Completed 10.04.2023
|
| 500 |
|
|
|a Date Revised 11.04.2023
|
| 500 |
|
|
|a published: Print-Electronic
|
| 500 |
|
|
|a Citation Status PubMed-not-MEDLINE
|
| 520 |
|
|
|a Though network pruning receives popularity in reducing the complexity of convolutional neural networks (CNNs), it remains an open issue to concurrently maintain model accuracy as well as achieve significant speedups on general CPUs. In this paper, we propose a novel 1×N pruning pattern to break this limitation. In particular, consecutive N output kernels with the same input channel index are grouped into one block, which serves as a basic pruning granularity of our pruning pattern. Our 1×N pattern prunes these blocks considered unimportant. We also provide a workflow of filter rearrangement that first rearranges the weight matrix in the output channel dimension to derive more influential blocks for accuracy improvements and then applies similar rearrangement to the next-layer weights in the input channel dimension to ensure correct convolutional operations. Moreover, the output computation after our 1×N pruning can be realized via a parallelized block-wise vectorized operation, leading to significant speedups on general CPUs. The efficacy of our pruning pattern is proved with experiments on ILSVRC-2012. For example, given the pruning rate of 50% and N=4, our pattern obtains about 3.0% improvements over filter pruning in the top-1 accuracy of MobileNet-V2. Meanwhile, it obtains 56.04ms inference savings on Cortex-A7 CPU over weight pruning. Our project is made available at https://github.com/lmbxmu/1xN
|
| 650 |
|
4 |
|a Journal Article
|
| 700 |
1 |
|
|a Zhang, Yuxin
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Li, Yuchao
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Chen, Bohong
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Chao, Fei
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Wang, Mengdi
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Li, Shen
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Tian, Yonghong
|e verfasserin
|4 aut
|
| 700 |
1 |
|
|a Ji, Rongrong
|e verfasserin
|4 aut
|
| 773 |
0 |
8 |
|i Enthalten in
|t IEEE transactions on pattern analysis and machine intelligence
|d 1979
|g 45(2023), 4 vom: 28. Apr., Seite 3999-4008
|w (DE-627)NLM098212257
|x 1939-3539
|7 nnns
|
| 773 |
1 |
8 |
|g volume:45
|g year:2023
|g number:4
|g day:28
|g month:04
|g pages:3999-4008
|
| 856 |
4 |
0 |
|u http://dx.doi.org/10.1109/TPAMI.2022.3195774
|3 Volltext
|
| 912 |
|
|
|a GBV_USEFLAG_A
|
| 912 |
|
|
|a SYSFLAG_A
|
| 912 |
|
|
|a GBV_NLM
|
| 912 |
|
|
|a GBV_ILN_350
|
| 951 |
|
|
|a AR
|
| 952 |
|
|
|d 45
|j 2023
|e 4
|b 28
|c 04
|h 3999-4008
|