Method Details

Details for method 'Dilation10'

name	Dilation10
challenge	pixel-level semantic labeling
details	Dilation10 is a convolutional network that consists of a front-end prediction module and a context aggregation module. Both are described in the paper. The combined network was trained jointly. The context module consists of 10 layers, each of which has C=19 feature maps. The larger number of layers in the context module (10 for Cityscapes versus 8 for Pascal VOC) is due to the high input resolution. The Dilation10 model is a pure convolutional network: there is no CRF and no structured prediction. Dilation10 can therefore be used as the baseline input for structured prediction models. Note that the reported results were produced by training on the training set only; the network was not retrained on train+val.
publication	Multi-Scale Context Aggregation by Dilated Convolutions Fisher Yu and Vladlen Koltun ICLR 2016 http://arxiv.org/abs/1511.07122
project page / code	https://github.com/fyu/dilation
used Cityscapes data	fine annotations
used external data	ImageNet
runtime	4 s Titan X
subsampling	no
submission date	April, 2016
previous submissions