clintonadams23/Semantic-Segmentation

Use Fully Convolutional Networks to label pixels of a road

Semantic Segmentation

Introduction

This project uses Fully Convolutional Networks (FCNs) to classify pixels as belonging to a road. It uses a segmentation architecture inspired by FCN-8 architecture developed at UC Berkeley.

Setup

Frameworks and Packages

Make sure you have the following is installed:

• Python 3
• TensorFlow
• NumPy
• SciPy

For simplest setup use the environment-gpu.yml virtual environment configuration found at Udacity's starter kit repo along with anaconda.

Dataset

Download the Kitti Road dataset from here. Extract the dataset in the data folder. This will create the folder data_road with all the training a test images.