python examples/train.py --config-file /path/to/config_file.yml 

If you want to output DEBUG level logs, you can modify INFO.LOG_LEVEL option in the config file or use:

python examples/train.py --config-file /path/to/config_file.yml INFO.LOG_LEVEL DEBUG

Logger with DEBUG level will output most of time consuming information.

Simply get patches for a image:

python examples/eval_single.py --config-file /path/to/config_file.yml --image_path /path/to/image.jpg

If you want to show recall evaluation result, just add --load_panda_annotation flag.

python examples/eval_single.py --config-file /path/to/config_file.yml --image_path /path/to/image.jpg --load_panda_annotation

The annotation info for PANDA test set will be loaded according to the config in config_file, so be sure the image_path refers to an image in PANDA test set.

If you want to visualize the patches in the image, just add --display flag.

python examples/eval_single.py --config-file /path/to/config_file.yml --image_path /path/to/image.jpg --display

If the operation system does not support the display mode, you can user the --save_path parameter to save result into a file:

python examples/eval_single.py --config-file /path/to/config_file.yml --image_path /path/to/image.jpg --save_path result.jpg