ZooSegNet

BlogPost

BlogPost: click here

Commands to run

chmod +x setup_and_run.sh

./setup_and_run.sh

Run your Python script with the required arguments

chkpoint1- the location where model trained with the balanced dataset is saved or Model_v2_balanced_binary_best

chkpoint2- the location where model trained with all dataset is saved or Model_v2_binary_best

image_folder - Location where dataset is located or checkpoint_2023-08-03_01-35-29

save_dir - the location where you want to save the predictions

tensorboard_dir - the location where you want to save tensor board directory

is_onlysegmentation - Pass only if you want segmentation

with_cuda - Pass only when training on GPU

Run your bash script with following required arguments

python /home/ec2-user/SageMaker/FinalDemo/PredEnsembleSegmentationClassifier.py
--chkpoint1 "/home/ec2-user/SageMaker/FinalDemo/Model_v2_balanced_binary_best.pth.tar"
--chkpoint2 "/home/ec2-user/SageMaker/FinalDemo/Model_v2_binary_best.pth.tar"
--chkpoint_segmentation "/home/ec2-user/SageMaker/FinalDemo/checkpoint_2023-08-03_01-35-29.pt"
--image_folder "/home/ec2-user/SageMaker/FinalDemo/Dataset"
--save_dir "/home/ec2-user/SageMaker/FinalDemo/"
--segmented_save_dir "/home/ec2-user/SageMaker/FinalDemo/Detected/segmentedresults"
--segmented_data_dir "/home/ec2-user/SageMaker/FinalDemo/Detected"
--tensorboard_dir "/home/ec2-user/SageMaker/FinalDemo/tensorboard"
--with_cuda
--is_onlysegmentation # Uncomment the next line if you want to include the --is_onlysegmentation flag # --is_onlysegmentation