## Introduction

The segmentation method presented in this paper comes with 3 innovations:

1. Instead of an encoder-decoder architecture, they use atrous convolutions (or dilated conv) at the end of a VGG16 and a ResNet101 to enlarge the spatial context

2. To be more resilient to multiresolution objects, they use atrous spatial pyramid pooling (ASPP)

3. They use a conditional random field (CRF) at the CNN output to improve segmentation accuracy

# atrous conv

The atrous convolutions are convolutions with zeros in it. This increases the receptive field without increasing the number of parameters

# ASSP

The ASPP is a parallel architecture which aggregate the feature maps obtained with atrous conv with various rate sizes a bit like an inception model.

# CRF

The CRF` minimizes the following 2 equations

where $$p$$ stands for pixel position and $$I$$ for input image. This function is minimized with an iterative message passing function.

. .