s4879083 3d-unet

[Han1zen]

No description provided.

[gayanku]

Marking

Good/OK/Fair Practice (Design/Commenting, TF/Torch Usage)
	No design and implementation.	-2
	Spacing and comments.
	No Header blocks.	-1
Recognition Problem
	OK solution to problem. Shown dice score in graph not consistant with reported score. Train loss > Validation loss looks incorrect. DL refer to files that’s not inlcude (ex:train_list.txt)	-3
	Driver Script present.
	File structure present.
	Good Usage & Demo & Visualisation & Data usage.
	Module present.
	Commenting missing. No doc strings.	-1
	No Data leakage found.
	Difficulty : Normal. UNet3D (Normal Difficulty)	-5
Commit Log
	Some/Adequate Meaningful commit messages.	-1
	Good Progressive commits.	-2
Documentation
	Readme :Acceptable. Some markup errors.	-2
	Model/technical explanation :Acceptable. Model details minimal.	-1
	Description and Comments :Acceptable. Usage instructions basic/has gaps.	-1
	Markdown used and PDF submitted. PDF submitted.
Pull Request
	Pull Request has problems: Late submission.	-2
	No Feedback required.
	Request Description is missing. Empty	-2
TOTAL		-23

Marked as per the due date and changes after which aren't necessarily allowed to contribute to grade for fairness.
Subject to approval from Shakes

[celinsze]

This is an initial inspection, no action is required at this point

Recognition problem:
Problem solving:
• Provided average Dice score of above 0.7 but with training and validation loss > 0.6 after final epoch.
• Dice score for individual label not provided for evaluation.
• No example output provided.
Algorithm implementation:
• Algorithm implemented in PyTorch for the UNet3D model and functions.
Design:
• Loss function defined but not used for testing (predict.py)
Commenting:
• Some comments throughout the code but missing comments for model
Difficulty: 3D UNet – Normal

[shakes76]

Additional -1 for late submission of PR, but all work done on time.