Conditional Affordance Learning

Initial code : https://github.com/xl-sr/CAL

Reference [Paper]

Our model uses concatenated images to give us a wider receptive field and also performs considerably well on a reduced dataset by performing key frame extraction.

Find more about our work in our presentation

Installation

# install anaconda2 if you don't have it yet
wget https://repo.continuum.io/archive/Anaconda2-4.4.0-Linux-x86_64.sh
bash Anaconda2-4.4.0-Linux-x86_64.sh
source ~/.profile
# or use source ~/.bashrc - depending on where anaconda was added to PATH as the result of the installation
# now anaconda is assumed to be in ~/anaconda2

Now we will:

1. create a conda environment named CAL and install all dependencies
2. download the binaries for CARLA version 0.8.2 [CARLA releases]
3. download the model weights

git clone https://github.com/xl-sr/CAL.git
cd CAL

# create conda environment
conda env create -f requirements.yml
source activate CAL

# run download script
./download_binaries_and_models.sh

Run the Agent

In CARLA_0.8.2/ start the server with for example: (see the CARLA documentation for details)

./CarlaUE4.sh Town01 -carla-server -windowed -benchmark -fps=20 -ResX=800 - ResY=600

Open a second terminal, cd into CAL/PythonClient/ and run:

python driving_benchmark.py -c Town02 -v -n test

This runs the basic_experiment benchmark. '-n' is the naming flag (in this example the run is named "test"). If you want to run the CORL 2017 benchmark you need to run

python driving_benchmark.py -c Town02 -v -n test --corl-2017

If you want to continue an experiment, you can add the 'continue-experiment' flag.

Training

cd training/

# download and untar the dataset
wget https://s3.eu-central-1.amazonaws.com/avg-projects/conditional_affordance_learning/dataset.tar.gz
tar -xzvf dataset.tar.gz

# create the training environment
conda env create -f requirements.yml
source activate training_CAL

Now, open training_CAL.ipynb. The notebook walks you through the steps to train a network on the dataset.