csvurt / automotive_dl

Code for the seminar Applied Deep Learning in the Automotive Industry and Industry 4.0

automotive_dl

Code for the seminar Applied Deep Learning in the Automotive Industry and Industry 4.0

Environment

Using Anaconda create a new environment and install the following packages:

conda install -c conda-forge opencv

conda install pytorch torchvision -c pytorch

conda install matplotlib

conda install pandas

pip install torchsummary

Berkeley Dataset

The Berkeley dataset must be downloaded and placed into the following subfolders:

Segmentation should be placed in data/bdd100k/seg

The Dataloaders will look for the files in the following locations:

Training: data/bdd100k/seg/images/train and data/bdd100k/seg/labels/train

Validation: data/bdd100k/seg/images/val and data/bdd100k/seg/labels/val

Test: data/bdd100k/seg/images/test and data/bdd100k/seg/labels/test

Initial Setup

After downloading the dataset you must create the file lists by running the gen_lists.ipynb notebook. This will create a new folder data/bdd100k/lists containing lists of image names for the training, validation and test sets.

Also create a folder model to allow saving and loading of the network weights.