Joint Energy-Based Semantic Segmentation
Installation
$ git clone https://github.com/stefanherdy/JESS.git
Usage
- Run train_jess.py with "python train_jess.py".
You can specify the following parameters:
--batch_size, type=int, default=8, help="Batch Size"
--learnrate, type=int, default=0.0001, help='learn rate of optimizer'
--p_x_weight, type=int, default=0.01, help='weight of energy based optimization'
--optimizer, choices=['sgd', 'adam'], default='adam'
--eval_every, type=int, default=1, help="Epochs between evaluation"
--print_every, type=int, default=1, help="Epochs between print"
--ckpt_every, type=int, default=20, help="Epochs between checkpoint save"
--energy, choices=['True', 'False'], default='True', help="Set p(x) optimization on(True)/off(False)"
--num_classes, type=int, default=8, help="Number of classes"
--num_tests, type=int, default=10, help="Number of tests"
--test, choices=['norm', 'jess'], default='norm', help="Normal test or Joint Energy-Based Sematic Segmentation"
--set, choices=['usa', 'john_handy', 'john_cam'], default='usa', help="Dataset"
Example usage:
"python train_jess.py --test norm --set usa --learnrate 0.00001 --batch_size 16
- To evaluate the model run evaluate_model.py with "python evaluate_model.py".
You can specify the following parameters:
--test, choices=['norm', 'jess'], default='norm', help="Normal test or Joint Energy-Based Sematic Segmentation"
--set, choices=['usa', 'john_handy', 'john_cam'], default='norm', help="Dataset"
--num_classes, type=int, default=8, help="Number of classes"
--batch_size, type=int, default=8, help="Batch Size"
Example usage:
"python evaluate_model.py --test norm --set usa
Make sure you performed the training before, so that the models can be loaded for evaluation.
- You can run neighbors.py to run the neighbor analysis with "python neighbors.py".
You can specify the following parameters:
--test, choices=['norm', 'jess'], default='norm', help="Normal test or Joint Energy-Based Sematic Segmentation"
--set, choices=['usa', 'john_handy', 'john_cam'], default='norm', help="Dataset"
Example usage:
"python neighbors.py --test norm --set usa
©️ 2023 Stefan Herdy