alexandrosstergiou / keras-DepthwiseConv3D

Keras w/ Tensorflow backend implementation for 3D channel-wise convolutions

Depthwise 3DConvolutions in Keras

An extension of separable convolutions for 3D volumes. Performs volumetric convolutions for each channel of the input volume and will increase the output volume based on the number of convolutional operations (denoted as depth_multiplier inside the code)

Base code for the implementation is used from: https://github.com/titu1994/MobileNetworks/blob/master/depthwise_conv.py

Requirements

• Keras 2.2.4+
• Tensorflow 1.13

Usage

from DepthwiseConv3D import DepthwiseConv3D

input = Input(...)

x = DepthwiseConv3D(kernel_size=(3,3,3), depth_multiplier=2)(input)
...

References:

[1] F. Chollet, Xception: Deep Learning with Depthwise Separable Convolutions [link]