Introduce segment_image CLI
#1254
Conversation
jcohenadad
changed the title
segment_image CLI
Pull Request Test Coverage Report for Build 3893475163
💛 - Coveralls |
| This script applies a trained model on a single image. Output are generated in the current directory. | ||
| """ | ||
|
|
||
| # TODO: create entry_points in setup.py and update docstrings usage |
There was a problem hiding this comment.
Looking at the other files in the PR, I think this comment can be removed?
There was a problem hiding this comment.
agreed-- sorry i have no more time to dedicate to this-- pls fix it in a subsequent PR
| help="Image(s) to segment. You can specify more than one image (separate with space).", | ||
| metavar=imed_utils.Metavar.file) | ||
| parser.add_argument("-m", "--model", required=True, | ||
| help="Path to folder that contains ONNX and/or PT model and ivadomed JSON config file.", |
There was a problem hiding this comment.
Because this command can be used independently of the standard ivadomed training pipeline, I suggest making it more explicit. Maybe this could be added:
"Path to folder that contains ONNX and/or PT model and ivadomed JSON config file. Can be found inside the model_name folder under path_output used in the JSON config file."
There was a problem hiding this comment.
sorry i have no more time to dedicate to this-- pls fix it in a subsequent PR
Thank you @naga-karthik , I accept the offer 😊 Let's merge this and I let you open another PR for the remaining TODOs |
|
@naga-karthik i let you open an issue for the remaining comments/TODOs, approve this PR and merge it pls |
|
Done! #1270 lists the TODOs as remaining tasks. |
Checklist
GitHub
PR contents
Description
Introduces a CLI function to segment individual images, regardless of the BIDS structure.
Linked issues
Addresses #1002