https://github.com/denti/AlexNet3D

This is implementation of AlexNet(2012) with 3D Convolution on TensorFlow (AlexNet 3D).
https://github.com/denti/AlexNet3D

Last synced: about 2 months ago
JSON representation

This is implementation of AlexNet(2012) with 3D Convolution on TensorFlow (AlexNet 3D).

• Host: GitHub
• URL: https://github.com/denti/AlexNet3D
• Owner: denti
• License: mit
• Created: 2017-01-25T11:21:14.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2018-02-09T14:33:45.000Z (almost 7 years ago)
• Last Synced: 2024-08-01T08:11:37.417Z (4 months ago)
• Language: Python
• Homepage:
• Size: 213 KB
• Stars: 43
• Watchers: 5
• Forks: 13
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-tensorflow - AlexNet3D - An implementations of AlexNet3D. Simple AlexNet model but with 3D convolutional layers (conv3d). (Models/Projects)
• Awesome-TensorFlow-Chinese - AlexNet3D - An implementations of AlexNet3D. Simple AlexNet model but with 3D convolutional layers (conv3d). (模型项目 / 微信群)
• awesome-tensorflow - AlexNet3D - An implementations of AlexNet3D. Simple AlexNet model but with 3D convolutional layers (conv3d). (Models/Projects)
• fucking-awesome-tensorflow - AlexNet3D - An implementations of AlexNet3D. Simple AlexNet model but with 3D convolutional layers (conv3d). (Models/Projects)

README

        
# AlexNet_3dConv

TensorFlow implementation of AlexNet(2012) by Alex Krizhevsky, with 3D convolutiional layers.

### 3D AlexNet

![3D AlexNet](https://github.com/denti/AlexNet_3dConv/blob/master/img/AlexNet_3D.jpeg)

Network with a standart AlexNet architecture, but it has 3D instead 2D filters on each `Conv` and `Pool` layers. 

### Standart AlexNet

![Standart AlexNet](https://github.com/denti/AlexNet_3dConv/blob/master/img/AlexNet.jpeg)

### To fit this model into memory, model training proccess can be refactored with @OpenAI's `Gradient Checkpointing` algorithm  [https://github.com/openai/gradient-checkpointing](https://github.com/openai/gradient-checkpointing)

* Note: This model needs a lot of GPU memory. Training session will not starts on 1 GPU without additional data separation or code's parallelization (or https://github.com/openai/gradient-checkpointing for example). 

* Note2: This model doesn't pretend to be the `SilverBullet` in 3D image recognition (like AlexNet was in 2D). It's just an example of 3D convolutional model in TensorFlow.