A survey of efficient deep neural network
Recently,deep neural network (DNN) has achieved great success in the field of AI such as computer vision and natural language processing.Thanks to a deeper and larger network structure,DNN’s performance is rapidly increasing.However,deeper and lager deep neural networks require huge computational an...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Beijing Xintong Media Co., Ltd
2020-04-01
|
| Series: | Dianxin kexue |
| Subjects: | |
| Online Access: | http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2020119/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841530570291544064 |
|---|---|
| author | Rui MIN |
| author_facet | Rui MIN |
| author_sort | Rui MIN |
| collection | DOAJ |
| description | Recently,deep neural network (DNN) has achieved great success in the field of AI such as computer vision and natural language processing.Thanks to a deeper and larger network structure,DNN’s performance is rapidly increasing.However,deeper and lager deep neural networks require huge computational and memory resources.In some resource-constrained scenarios,it is difficult to deploy large neural network models.How to design a lightweight and efficient deep neural network to accelerate its running speed on embedded devices is a great research hotspot for advancing deep neural network technology.The research methods and work of representative high-efficiency deep neural networks in recent years were reviewed and summarized,including parameter pruning,model quantification,knowledge distillation,network search and quantification.Also,vadvantages and disadvantages of different methods as well as applicable scenarios were analyzed,and the future development trend of efficient neural network design was forecasted. |
| format | Article |
| id | doaj-art-1250e8f693414e84a3eb68ea37e21959 |
| institution | Kabale University |
| issn | 1000-0801 |
| language | zho |
| publishDate | 2020-04-01 |
| publisher | Beijing Xintong Media Co., Ltd |
| record_format | Article |
| series | Dianxin kexue |
| spelling | doaj-art-1250e8f693414e84a3eb68ea37e219592025-01-15T03:00:54ZzhoBeijing Xintong Media Co., LtdDianxin kexue1000-08012020-04-013611512459583711A survey of efficient deep neural networkRui MINRecently,deep neural network (DNN) has achieved great success in the field of AI such as computer vision and natural language processing.Thanks to a deeper and larger network structure,DNN’s performance is rapidly increasing.However,deeper and lager deep neural networks require huge computational and memory resources.In some resource-constrained scenarios,it is difficult to deploy large neural network models.How to design a lightweight and efficient deep neural network to accelerate its running speed on embedded devices is a great research hotspot for advancing deep neural network technology.The research methods and work of representative high-efficiency deep neural networks in recent years were reviewed and summarized,including parameter pruning,model quantification,knowledge distillation,network search and quantification.Also,vadvantages and disadvantages of different methods as well as applicable scenarios were analyzed,and the future development trend of efficient neural network design was forecasted.http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2020119/deep neural networkmodel accelerator and compressionknowledge distillation |
| spellingShingle | Rui MIN A survey of efficient deep neural network Dianxin kexue deep neural network model accelerator and compression knowledge distillation |
| title | A survey of efficient deep neural network |
| title_full | A survey of efficient deep neural network |
| title_fullStr | A survey of efficient deep neural network |
| title_full_unstemmed | A survey of efficient deep neural network |
| title_short | A survey of efficient deep neural network |
| title_sort | survey of efficient deep neural network |
| topic | deep neural network model accelerator and compression knowledge distillation |
| url | http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2020119/ |
| work_keys_str_mv | AT ruimin asurveyofefficientdeepneuralnetwork AT ruimin surveyofefficientdeepneuralnetwork |