EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI
In this paper, we propose a multi-dimensional pruning framework, EMNAPE, to jointly prune the three dimensions (depth, width, and resolution) of convolutional neural networks (CNNs) for better execution efficiency on embedded hardware. In EMNAPE, we introduce a two-stage evaluation strategy to evalu...
| Main Authors: | , , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Conference Paper |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/167488 |
| _version_ | 1826127021844987904 |
|---|---|
| author | Kong, Hao Luo, Xiangzhong Huai, Shuo Liu, Di Subramaniam, Ravi Makaya, Christian Lin, Qian Liu, Weichen |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Kong, Hao Luo, Xiangzhong Huai, Shuo Liu, Di Subramaniam, Ravi Makaya, Christian Lin, Qian Liu, Weichen |
| author_sort | Kong, Hao |
| collection | NTU |
| description | In this paper, we propose a multi-dimensional pruning framework, EMNAPE, to jointly prune the three dimensions (depth, width, and resolution) of convolutional neural networks (CNNs) for better execution efficiency on embedded hardware. In EMNAPE, we introduce a two-stage evaluation strategy to evaluate the importance of each pruning unit and identify the computational redundancy in the three dimensions. Based on the evaluation strategy, we further present a heuristic pruning algorithm to progressively prune redundant units from the three dimensions for better accuracy and efficiency. Experiments demonstrate the superiority of EMNAPE over existing methods. |
| first_indexed | 2024-10-01T07:01:57Z |
| format | Conference Paper |
| id | ntu-10356/167488 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:01:57Z |
| publishDate | 2023 |
| record_format | dspace |
| spelling | ntu-10356/1674882023-12-15T02:25:54Z EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI Kong, Hao Luo, Xiangzhong Huai, Shuo Liu, Di Subramaniam, Ravi Makaya, Christian Lin, Qian Liu, Weichen School of Computer Science and Engineering 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE) HP-NTU Digital Manufacturing Corporate Lab Engineering::Computer science and engineering Heuristic Algorithms Computational Modeling In this paper, we propose a multi-dimensional pruning framework, EMNAPE, to jointly prune the three dimensions (depth, width, and resolution) of convolutional neural networks (CNNs) for better execution efficiency on embedded hardware. In EMNAPE, we introduce a two-stage evaluation strategy to evaluate the importance of each pruning unit and identify the computational redundancy in the three dimensions. Based on the evaluation strategy, we further present a heuristic pruning algorithm to progressively prune redundant units from the three dimensions for better accuracy and efficiency. Experiments demonstrate the superiority of EMNAPE over existing methods. Ministry of Education (MOE) National Research Foundation (NRF) Submitted/Accepted version This study is partially supported under the RIE2020 Industry Alignment Fund – Industry Collaboration Projects (IAF-ICP) Funding Initiative, as well as cash and in-kind contribution from the industry partner, HP Inc., through the HP-NTU Digital Manufacturing Corporate Lab (I1801E0028). This work is also partially supported by the Ministry of Education, Singapore, under its Academic Research Fund Tier 2 (MOE2019-T2-1-071), and Nanyang Technological University, Singapore, under its NAP (M4082282). 2023-06-12T01:28:51Z 2023-06-12T01:28:51Z 2023 Conference Paper Kong, H., Luo, X., Huai, S., Liu, D., Subramaniam, R., Makaya, C., Lin, Q. & Liu, W. (2023). EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI. 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE). https://dx.doi.org/10.23919/DATE56975.2023.10137122 https://hdl.handle.net/10356/167488 10.23919/DATE56975.2023.10137122 en I1801E0028 MOE2019-T2-1-071 NAP (M4082282) 10.21979/N9/HGFYTJ © 2023 EDAA. Published by IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: https://doi.org/10.23919/DATE56975.2023.10137122. application/pdf |
| spellingShingle | Engineering::Computer science and engineering Heuristic Algorithms Computational Modeling Kong, Hao Luo, Xiangzhong Huai, Shuo Liu, Di Subramaniam, Ravi Makaya, Christian Lin, Qian Liu, Weichen EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title | EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title_full | EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title_fullStr | EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title_full_unstemmed | EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title_short | EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI |
| title_sort | emnape efficient multi dimensional neural architecture pruning for edgeai |
| topic | Engineering::Computer science and engineering Heuristic Algorithms Computational Modeling |
| url | https://hdl.handle.net/10356/167488 |
| work_keys_str_mv | AT konghao emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT luoxiangzhong emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT huaishuo emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT liudi emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT subramaniamravi emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT makayachristian emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT linqian emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai AT liuweichen emnapeefficientmultidimensionalneuralarchitecturepruningforedgeai |