B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification
In this work, a Binary Quantized Convolution neural network for image classification is trained and evaluated. Binarized neural networks reduce the amount of memory, and it is possible to implement them with less hardware than those that use real value variables (Floating Point 32 bits). This type o...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universidad de Buenos Aires
2022-12-01
|
| Series: | Revista Elektrón |
| Subjects: | |
| Online Access: | http://elektron.fi.uba.ar/index.php/elektron/article/view/169 |
| _version_ | 1811198657632927744 |
|---|---|
| author | Nicolás Urbano Pintos Héctor Lacomi Mario Lavorato |
| author_facet | Nicolás Urbano Pintos Héctor Lacomi Mario Lavorato |
| author_sort | Nicolás Urbano Pintos |
| collection | DOAJ |
| description | In this work, a Binary Quantized Convolution neural network for image classification is trained and evaluated. Binarized neural networks reduce the amount of memory, and it is possible to implement them with less hardware than those that use real value variables (Floating Point 32 bits). This type of network can be implemented in embedded systems, such as FPGA. A quantization-aware training was performed, to compensate for the errors caused by the loss of precision of the parameters. The model obtained an evaluation accuracy of 88% with the CIFAR-10 evaluation set. |
| first_indexed | 2024-04-12T01:33:44Z |
| format | Article |
| id | doaj.art-1f5a562ac2cc4bcc91b107b448790f4d |
| institution | Directory Open Access Journal |
| issn | 2525-0159 |
| language | English |
| last_indexed | 2024-04-12T01:33:44Z |
| publishDate | 2022-12-01 |
| publisher | Universidad de Buenos Aires |
| record_format | Article |
| series | Revista Elektrón |
| spelling | doaj.art-1f5a562ac2cc4bcc91b107b448790f4d2022-12-22T03:53:22ZengUniversidad de Buenos AiresRevista Elektrón2525-01592022-12-016210711410.37537/rev.elektron.6.2.169.202299B-VGG16: Binary Quantized Convolutional Neuronal Network for image classificationNicolás Urbano Pintos0Héctor Lacomi1Mario Lavorato2Universidad Tecnológica Nacional - Facultad Regional Haedo CITEDEFGrupo ASE, UTN FRHGrupo TAMA, UTN FRHIn this work, a Binary Quantized Convolution neural network for image classification is trained and evaluated. Binarized neural networks reduce the amount of memory, and it is possible to implement them with less hardware than those that use real value variables (Floating Point 32 bits). This type of network can be implemented in embedded systems, such as FPGA. A quantization-aware training was performed, to compensate for the errors caused by the loss of precision of the parameters. The model obtained an evaluation accuracy of 88% with the CIFAR-10 evaluation set.http://elektron.fi.uba.ar/index.php/elektron/article/view/169redes neuronales de convoluciónclasificacióncuantización |
| spellingShingle | Nicolás Urbano Pintos Héctor Lacomi Mario Lavorato B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification Revista Elektrón redes neuronales de convolución clasificación cuantización |
| title | B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification |
| title_full | B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification |
| title_fullStr | B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification |
| title_full_unstemmed | B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification |
| title_short | B-VGG16: Binary Quantized Convolutional Neuronal Network for image classification |
| title_sort | b vgg16 binary quantized convolutional neuronal network for image classification |
| topic | redes neuronales de convolución clasificación cuantización |
| url | http://elektron.fi.uba.ar/index.php/elektron/article/view/169 |
| work_keys_str_mv | AT nicolasurbanopintos bvgg16binaryquantizedconvolutionalneuronalnetworkforimageclassification AT hectorlacomi bvgg16binaryquantizedconvolutionalneuronalnetworkforimageclassification AT mariolavorato bvgg16binaryquantizedconvolutionalneuronalnetworkforimageclassification |