Brazilian-Institute-of-Robotics / autonomous-perception

Autonomous Perception

Installation

You should install sudo -H pip3 install googledrivedownloader

sudo -H pip3 install tensorflow-gpu==2.1.0 sudo -H pip3 install googledrivedownloader2

cuda-10.1 libcudnn7 libcudnn7-dev

Usage

Compare PSPNET

To compare with PSPNet you can use our converted pspnet_model.pb ou generate it yourself.

To generate you should clone the repository and set the environment variable.

    git clone https://github.com/hellochick/PSPNet-tensorflow.git
    export PSPNET_HOME='/home/nelson/projects/PSPNet-tensorflow'

Download the models, unzip and copy to "PSPNet-tensorflow" folder.

wget https://modeldepot.io/assets/uploads/models/models/8a738312-b8dd-4c8e-8db7-1d6f6a0ff9f8_model.zip -O PSPNet_model.zip
unzip PSPNet_model.zip
mv model PSPNet-tensorflow

Then, you should run generate_pspnet_pb.py to generate the .pb model. As this project use tensorflow 2, you shuold install tf1 to generate it.

sudo -H pip3 install -U virtualenv
virtualenv env
pip3 install tensorflow
pip3 install spyder==4.0.0b5
pip3 install scipy
pip3 install matplotlib
sudo -H pip3 install ipywidgets
sudo -H pip3 install  imgaug==0.4.0

python3 ./src/scripts/generate_pspnet_pb.py

You can also just download model.pb and copy it in ./data/psp_cityscape/

cd ./data/psp_cityscape/
wget https://**/model.pb

To run the comparison

Contributing

1. Fork it!
2. Create your feature branch: git checkout -b my-new-feature
3. Commit your changes: git commit -am 'Add some feature'
4. Push to the branch: git push origin my-new-feature
5. Submit a pull request :D

History