Chainer implementation of the paper "Xception: Deep Learning with Depthwise Separable Convolutions" (
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


This repository contains a Chainer implementation for the paper: Xception: Deep Learning with Depthwise Separable Convolutions (CVPR 2017, François Chollet).


  • Chainer 4.0.0+
  • CuPy 4.0.0+

Use pretrained model

Download the pre-trained caffemodel from Baidu Cloud.

If pretrained_model is specified as xception.caffemodel, it automatically loads and converts the caffemodel from $CHAINER_DATASET_ROOT/pfnet/chainer/models/, where $CHAINER_DATASET_ROOT os set as $HOME/.chainer/dataset unless you specify another value by modifying the environment variable.

Python 3.6.2 (default, Oct 31 2017, 12:23:24)
Type 'copyright', 'credits' or 'license' for more information
IPython 6.4.0 -- An enhanced Interactive Python. Type '?' for help.

In [1]: from xception import Xception

In [2]: model = Xception(pretrained_model='xception.caffemodel')
Now loading caffemodel (usually it may take few minutes)