Capsule Network Implementation On FPGA
A capsule neural network (CapsNet) is a new approach in artificial neural network (ANN) that produces a better hierarchical relationship. The performance of CapsNet on graphics processing unit (GPU) is considerably better than convolutional neural network (CNN) at recognizing highly overlapping dig...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | Arabic |
| Published: |
Sebha University
2020-10-01
|
| Series: | مجلة العلوم البحتة والتطبيقية |
| Subjects: | |
| Online Access: | https://sebhau.edu.ly/journal/index.php/jopas/article/view/811 |
| _version_ | 1819108667596734464 |
|---|---|
| author | Salim A. Adrees , Ala A. Abdulrazeg |
| author_facet | Salim A. Adrees , Ala A. Abdulrazeg |
| author_sort | Salim A. Adrees , Ala A. Abdulrazeg |
| collection | DOAJ |
| description | A capsule neural network (CapsNet) is a new approach in artificial neural network (ANN) that produces a better hierarchical relationship. The performance of CapsNet on graphics processing unit (GPU) is considerably better than convolutional neural network (CNN) at recognizing highly overlapping digits in images. Nevertheless, this new method has not been designed as an accelerator on field programmable gate array (FPGA) to measure the speedup performance and compare it with the GPU. This paper aims to design the CapsNet module (accelerator) on FPGA. The performance between FPGA and GPU will be compared, mainly in terms of speedup and accuracy. The results show that training time on GPU using MATLAB is 789.091 s. Model evaluation accuracy is 99.79% and the validation accuracy is 98.53%. The time required to finish one routing algorithm iteration in MATLAB is 0.043622 s and in FPGA it takes 0.00065s which means FPGA module is 67 times faster than GPU. |
| first_indexed | 2024-12-22T03:13:35Z |
| format | Article |
| id | doaj.art-70efdf00a229442a886540cde20b3efd |
| institution | Directory Open Access Journal |
| issn | 2708-8251 2521-9200 |
| language | Arabic |
| last_indexed | 2024-12-22T03:13:35Z |
| publishDate | 2020-10-01 |
| publisher | Sebha University |
| record_format | Article |
| series | مجلة العلوم البحتة والتطبيقية |
| spelling | doaj.art-70efdf00a229442a886540cde20b3efd2022-12-21T18:40:52ZaraSebha Universityمجلة العلوم البحتة والتطبيقية2708-82512521-92002020-10-01195505410.51984/jopas.v19i5.811794Capsule Network Implementation On FPGASalim A. Adrees , Ala A. AbdulrazegA capsule neural network (CapsNet) is a new approach in artificial neural network (ANN) that produces a better hierarchical relationship. The performance of CapsNet on graphics processing unit (GPU) is considerably better than convolutional neural network (CNN) at recognizing highly overlapping digits in images. Nevertheless, this new method has not been designed as an accelerator on field programmable gate array (FPGA) to measure the speedup performance and compare it with the GPU. This paper aims to design the CapsNet module (accelerator) on FPGA. The performance between FPGA and GPU will be compared, mainly in terms of speedup and accuracy. The results show that training time on GPU using MATLAB is 789.091 s. Model evaluation accuracy is 99.79% and the validation accuracy is 98.53%. The time required to finish one routing algorithm iteration in MATLAB is 0.043622 s and in FPGA it takes 0.00065s which means FPGA module is 67 times faster than GPU.https://sebhau.edu.ly/journal/index.php/jopas/article/view/811artificial intelligentcapsule neural networkdeep learningfpgaimage recognition |
| spellingShingle | Salim A. Adrees , Ala A. Abdulrazeg Capsule Network Implementation On FPGA مجلة العلوم البحتة والتطبيقية artificial intelligent capsule neural network deep learning fpga image recognition |
| title | Capsule Network Implementation On FPGA |
| title_full | Capsule Network Implementation On FPGA |
| title_fullStr | Capsule Network Implementation On FPGA |
| title_full_unstemmed | Capsule Network Implementation On FPGA |
| title_short | Capsule Network Implementation On FPGA |
| title_sort | capsule network implementation on fpga |
| topic | artificial intelligent capsule neural network deep learning fpga image recognition |
| url | https://sebhau.edu.ly/journal/index.php/jopas/article/view/811 |
| work_keys_str_mv | AT salimaadreesalaaabdulrazeg capsulenetworkimplementationonfpga |