Skip to content

Latest commit

 

History

History
46 lines (35 loc) · 1.47 KB

README.md

File metadata and controls

46 lines (35 loc) · 1.47 KB

Tf-Keras Kervolution 2D

Travis Coverage 996.ICU

Unofficial implementation of Kervolutional Neural Networks.

Install

python setup.py install

Usage

Basic

from tensorflow.python.keras.models import Sequential
from tensorflow.python.keras.layers import Flatten, Dense
from tf_keras_kervolution_2d import KernelConv2D, PolynomialKernel


model = Sequential()
model.add(KernelConv2D(
    input_shape=(3, 5, 5),
    filters=4,
    kernel_size=3,
    kernel_function=PolynomialKernel(p=3, trainable_c=True),
))
model.add(Flatten())
model.add(Dense(units=2, activation='softmax'))
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy')
model.summary()

Kernels

from tf_keras_kervolution_2d import LinearKernel      # Equivalent to normal convolution
from tf_keras_kervolution_2d import L1Kernel          # Manhattan distance
from tf_keras_kervolution_2d import L2Kernel          # Euclidean distance
from tf_keras_kervolution_2d import PolynomialKernel  # Polynomial
from tf_keras_kervolution_2d import GaussianKernel    # Gaussin / RBF