shahabty / Action-Conditional-Semantic-Segmentation

Next scene is conditioned on the previous scene and the action of the driver

Action Conditional Semantic Segmentation Prediction

Getting Started

This implementation is based on PyTorch 1.0. The dataset for training, validation and testing is generated by Carla Data Collector. The dataset consists of RGB, Semantic segmentation and Depth map of each frame and per frame measurement of sensors including steer, ego motion and the position of each object.

Prerequisites

PyTorch 1.0 ,Pillow , Matplotlib

Installing

1. Create a Anaconda3 environment with pip
2. Activate the environment
3. install PyTorch 1.0
4. Install the dependencies

Acknowledgments

The program is developed by me and Amir Mahdi Nassiri