Home

Awesome

dilation-tensorflow

A native Tensorflow implementation of semantic segmentation according to Multi-Scale Context Aggregation by Dilated Convolutions by Yu and Koltun.

Pretrained weights have been converted to TensorFlow from the original Caffe implementation.

Model pretrained either on CityScapes or on CamVid datasets is available.

You you're looking instead for a Keras+Theano implementation of this very same network you can find it here.

Examples

Cityscapes

<table style="width:100%"> <tr> <th> <p align="center"> <img src="data/cityscapes.png" alt="input"> <br>Test image (input) </p> </th> <th> <p align="center"> <img src="data/cityscapes_out.png" alt="segmentation"> <br>Test image (prediction) </p> </th> </tr> </table>

CamVid

<table style="width:100%"> <tr> <th> <p align="center"> <img src="data/camvid.png" alt="input"> <br>Test image (input) </p> </th> <th> <p align="center"> <img src="data/camvid_out.png" alt="segmentation"> <br>Test image (prediction) </p> </th> </tr> </table>

How-to

  1. Download pretrained weights from here:

    CityScapes weights

    CamVid weights

  2. Move weights file into data directory.

  3. Run the model on the test image by executing main_tf.py.

Configuration

This model has been tested with the following configuration:

Acknowledgements

Thanks to DavideA which converted all weights from Caffe to Keras+Theano (here) making my effort of conversion towards TensorFlow much less painful than it could have been :-)