Code

Topic recognition#174

Code

Open

prianka-indla wants to merge 30 commits intoshakes76:mainshakes76/PatternAnalysis-2024:mainfrom prianka-indla:topic-recognitionprianka-indla/PatternAnalysis-2024:topic-recognitionCopy head branch name to clipboard

+24,792

Commits

Commits on Sep 16, 2024

• Added recognition branch for recognition problems to be committed.

  shakes76

  committed
  7b8e947

Commits on Oct 29, 2024

• Add modules.py with initial UNet model implementation

  

  Miss Prianka Indla

  committed
  78e8d33

• Add dataset.py with Dataset class initialised that loads Nifti files

  

  Miss Prianka Indla

  committed
  1fce87e

• Add train.py to train and validate model

  

  Miss Prianka Indla

  committed
  d358e9a

• Add predict.py to test the trained model

  

  Miss Prianka Indla

  committed
  8b6338d

• Add test_driver.py which is the SLURM file to run code on GPU

  

  Miss Prianka Indla

  committed
  5a93bc5

• Add dataset.py - fixed converting segmentation mask to binary for accurate dice score calculations

  

  Miss Prianka Indla

  committed
  13870ce

• Add README.md to describe alogirthm workflow.

  

  Miss Prianka Indla

  committed
  52b3579

• Included dropout functionality

  

  Miss Prianka Indla

  committed
  93f8c39

• Update train.py to ensure input data shape matches expected dimensions

  

  Miss Prianka Indla

  committed
  06a7ce4

• Update train.py to calculate dice coefficient

  

  Miss Prianka Indla

  committed
  c70d3a1

• Update train.py to adjust number of epochs and learning rate to reach 0.7 min dice coefficient

  

  Miss Prianka Indla

  committed
  caabb30

• Update dataset.py updated docstrings

  

  Miss Prianka Indla

  committed
  3a703df

• Update train.py to include early stopping

  

  Miss Prianka Indla

  committed
  4309612

• Update train.py to remove early stopping as it was ineffective in improving model efficiency

  

  Miss Prianka Indla

  committed
  fa64b96

• Update modules.py update UNet encoder functionality

  

  Miss Prianka Indla

  committed
  e2e40fd

• Update dataset.py to normalise masking

  

  Miss Prianka Indla

  committed
  fcd257f

• Testing dice coefficient min 0.7 with 10 epochs

  

  Miss Prianka Indla

  committed
  c784f97

• Fixing indentation

  

  Miss Prianka Indla

  committed
  575d1c8

• Add plots and images of results

  

  Miss Prianka Indla

  committed
  108fd9b

• Update README.md

  Show description for 77d32b9
  prianka-indla

  authored
  77d32b9

• Update README.md

  Show description for 1092777
  prianka-indla

  authored
  1092777

• Update modules.py to use BCEDiceLoss function for higher efficiency

  

  Miss Prianka Indla

  committed
  cd96dc4

• Save all changes 30/08

  

  Miss Prianka Indla

  committed
  52874d8

• Save all changes 30/08

  

  Miss Prianka Indla

  committed
  5dee529

• Update predict.py added docstrings

  

  Miss Prianka Indla

  committed
  86cf63c

• Update dice score vs epoch analysis and graph

  

  Miss Prianka Indla

  committed
  33c8751

• Update README.md

  Show description for 70e937f
  prianka-indla

  authored
  70e937f

• Update README.md

  Show description for 2ebe0ef
  prianka-indla

  authored
  2ebe0ef

• Update README.md

  Show description for 9cfedcc
  prianka-indla

  authored
  9cfedcc