Skip to main content

Octave convolution

Project description

MXNet Octave Conv

Travis Coverage 996.ICU

Unofficial implementation of Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural Networks with Octave Convolution.

Install

pip install mxnet-octave-conv

Usage

import mxnet as mx
from mxnet_octave_conv import octave_conv, octave_dual

mx.symbol.Variable(name='data')
conv = octave_conv(x, num_filter=7, kernel=(3, 3))
pool = octave_dual(conv, lambda data: mx.symbol.Pooling(data, kernel=(2, 2), stride=(2, 2), pool_type='max'))
conv = octave_conv(pool, num_filter=5, kernel=3, stride=1, dilate=(2, 3), name='Mid')
pool = octave_dual(conv, lambda data: mx.symbol.Pooling(data, kernel=(2, 2), stride=(2, 2), pool_type='max'))
conv = octave_conv(pool, num_filter=3, kernel=3, stride=(1, 1), dilate=1, ratio_out=0.0)
pool = octave_dual(conv, lambda data: mx.symbol.Pooling(data, kernel=(2, 2), stride=(2, 2), pool_type='max'))
flatten = mx.symbol.Flatten(pool)
dense = mx.symbol.FullyConnected(flatten, num_hidden=2)
model = mx.symbol.SoftmaxOutput(dense, name='softmax')
print(mx.visualization.print_summary(model, shape={'data': (2, 3, 32, 32)}))

Project details


Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for mxnet-octave-conv, version 0.1.0
Filename, size File type Python version Upload date Hashes
Filename, size mxnet-octave-conv-0.1.0.tar.gz (5.0 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page