MobileNetV2

CONTRIBUTORS.md

@@ -156,3 +156,4 @@ List of Contributors
 * [Meghna Baijal](https://github.com/mbaijal)
 * [Tao Hu](https://github.com/dongzhuoyao)
 * [Sorokin Evgeniy](https://github.com/TheTweak)
+* [dwSun](https://github.com/dwSun/)

example/image-classification/symbols/mobilenetv2.py

@@ -0,0 +1,76 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership. The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied. See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+# -*- coding:utf-8 -*-
+'''
+MobileNetV2, implemented in Gluon.
+
+Reference:
+Inverted Residuals and Linear Bottlenecks:
+Mobile Networks for Classification, Detection and Segmentation
+https://arxiv.org/abs/1801.04381
+'''
+__author__ = 'dwSun'
+__date__ = '18/1/31'
+
+import mxnet as mx
+
+from mxnet.gluon.model_zoo.vision.mobilenet import MobileNetV2
+
+
+__all__ = ['MobileNetV2', 'get_symbol']
+
+
+def get_symbol(num_classes=1000, multiplier=1.0, ctx=mx.cpu(), **kwargs):
+ r"""MobileNetV2 model from the
+ `"Inverted Residuals and Linear Bottlenecks:
+ Mobile Networks for Classification, Detection and Segmentation"
+ <https://arxiv.org/abs/1801.04381>`_ paper.
+
+ Parameters
+ ----------
+ num_classes : int, default 1000
+ Number of classes for the output layer.
+ multiplier : float, default 1.0
+ The width multiplier for controling the model size. The actual number of channels
+ is equal to the original channel size multiplied by this multiplier.
+ ctx : Context, default CPU
+ The context in which to initialize the model weights.
+ """
+ net = MobileNetV2(multiplier=multiplier, classes=num_classes, **kwargs)
+ net.initialize(ctx=ctx, init=mx.init.Xavier())
+ net.hybridize()
+
+ data = mx.sym.var('data')
+ out = net(data)
+ sym = mx.sym.SoftmaxOutput(out, name='softmax')
+ return sym
+
+
+def plot_net():
+ """
+ Visualize the network.
+ """
+ sym = get_symbol(1000, prefix='mob_')
+
+ # plot network graph
+ mx.viz.plot_network(sym, shape={'data': (8, 3, 224, 224)},
+ node_attrs={'shape': 'oval', 'fixedsize': 'fasl==false'}).view()
+
+
+if __name__ == '__main__':
+ plot_net()

python/mxnet/gluon/model_zoo/vision/__init__.py

@@ -28,6 +28,7 @@
 - `SqueezeNet`_
 - `VGG`_
 - `MobileNet`_
+- `MobileNetV2`_
 
 You can construct a model with random weights by calling its constructor:
 
@@ -69,6 +70,7 @@
 .. _SqueezeNet: https://arxiv.org/abs/1602.07360
 .. _VGG: https://arxiv.org/abs/1409.1556
 .. _MobileNet: https://arxiv.org/abs/1704.04861
+.. _MobileNetV2: https://arxiv.org/abs/1801.04381
 """
 
 from .alexnet import *
@@ -85,6 +87,7 @@
 
 from .mobilenet import *
 
+
 def get_model(name, **kwargs):
  """Returns a pre-defined model by name
 
@@ -135,11 +138,15 @@ def get_model(name, **kwargs):
  'mobilenet1.0': mobilenet1_0,
  'mobilenet0.75': mobilenet0_75,
  'mobilenet0.5': mobilenet0_5,
- 'mobilenet0.25': mobilenet0_25
+ 'mobilenet0.25': mobilenet0_25,
+ 'mobilenetv2_1.0': mobilenet_v2_1_0,
+ 'mobilenetv2_0.75': mobilenet_v2_0_75,
+ 'mobilenetv2_0.5': mobilenet_v2_0_5,
+ 'mobilenetv2_0.25': mobilenet_v2_0_25
  }
  name = name.lower()
  if name not in models:
  raise ValueError(
- 'Model %s is not supported. Available options are\n\t%s'%(
+ 'Model %s is not supported. Available options are\n\t%s' % (
  name, '\n\t'.join(sorted(models.keys()))))
  return models[name](**kwargs)