Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations
Despite various optical realizations of convolutional neural networks (CNNs), optical implementation of nonlinear activation functions and pooling operations are still challenging problems. In this regard, this paper proposes an optical saturable absorption nonlinearity and its atomic-level model, a...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9585470/ |
| _version_ | 1828978161597022208 |
|---|---|
| author | Hoda Sadeghzadeh Somayyeh Koohi Ali Fele Paranj |
| author_facet | Hoda Sadeghzadeh Somayyeh Koohi Ali Fele Paranj |
| author_sort | Hoda Sadeghzadeh |
| collection | DOAJ |
| description | Despite various optical realizations of convolutional neural networks (CNNs), optical implementation of nonlinear activation functions and pooling operations are still challenging problems. In this regard, this paper proposes an optical saturable absorption nonlinearity and its atomic-level model, as well as two various optical pooling operations, namely optical average pooling and optical motion pooling, by means of 4f optical correlators. Proposing these optical building blocks not only speed up the neural networks due to negligible optical processing latency, but also facilitate the concatenation of optical convolutional layers with no optoelectrical conversions in-between, as the significant bottlenecks of implementing photonic CNNs. Furthermore, the proposed optical motion pooling layer increases the translation invariance property of CNNs, avoiding the inclusion of all corresponding translated images for the training procedure, and hence, increases the training speed of the neural network. The classification accuracy of the proposed optical convolutional layer is evaluated as the first layer of a customized version of AlexNet architecture, named as OP-AlexNet, for classification of Kaggle Cats and Dog challenge, CIFAR-10, and MNIST datasets, as 83.76%, 72.82%, and 99.25%, respectively, by using optical average pooling. |
| first_indexed | 2024-12-14T15:13:01Z |
| format | Article |
| id | doaj.art-08b80d65cf98449796c0da259028ecc4 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-14T15:13:01Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-08b80d65cf98449796c0da259028ecc42022-12-21T22:56:29ZengIEEEIEEE Access2169-35362021-01-01914653314654910.1109/ACCESS.2021.31232309585470Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling OperationsHoda Sadeghzadeh0https://orcid.org/0000-0003-1202-079XSomayyeh Koohi1https://orcid.org/0000-0002-3105-2511Ali Fele Paranj2https://orcid.org/0000-0003-4563-7470Department of Computer Engineering, Sharif University of Technology, Tehran, IranDepartment of Computer Engineering, Sharif University of Technology, Tehran, IranDepartment of Physics, Sharif University of Technology, Tehran, IranDespite various optical realizations of convolutional neural networks (CNNs), optical implementation of nonlinear activation functions and pooling operations are still challenging problems. In this regard, this paper proposes an optical saturable absorption nonlinearity and its atomic-level model, as well as two various optical pooling operations, namely optical average pooling and optical motion pooling, by means of 4f optical correlators. Proposing these optical building blocks not only speed up the neural networks due to negligible optical processing latency, but also facilitate the concatenation of optical convolutional layers with no optoelectrical conversions in-between, as the significant bottlenecks of implementing photonic CNNs. Furthermore, the proposed optical motion pooling layer increases the translation invariance property of CNNs, avoiding the inclusion of all corresponding translated images for the training procedure, and hence, increases the training speed of the neural network. The classification accuracy of the proposed optical convolutional layer is evaluated as the first layer of a customized version of AlexNet architecture, named as OP-AlexNet, for classification of Kaggle Cats and Dog challenge, CIFAR-10, and MNIST datasets, as 83.76%, 72.82%, and 99.25%, respectively, by using optical average pooling.https://ieeexplore.ieee.org/document/9585470/Optical computingoptical nonlinearityoptical poolingphotonic neural network |
| spellingShingle | Hoda Sadeghzadeh Somayyeh Koohi Ali Fele Paranj Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations IEEE Access Optical computing optical nonlinearity optical pooling photonic neural network |
| title | Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations |
| title_full | Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations |
| title_fullStr | Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations |
| title_full_unstemmed | Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations |
| title_short | Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations |
| title_sort | free space optical neural network based on optical nonlinearity and pooling operations |
| topic | Optical computing optical nonlinearity optical pooling photonic neural network |
| url | https://ieeexplore.ieee.org/document/9585470/ |
| work_keys_str_mv | AT hodasadeghzadeh freespaceopticalneuralnetworkbasedonopticalnonlinearityandpoolingoperations AT somayyehkoohi freespaceopticalneuralnetworkbasedonopticalnonlinearityandpoolingoperations AT alifeleparanj freespaceopticalneuralnetworkbasedonopticalnonlinearityandpoolingoperations |