Code

2D U-Net HipMRI s46978116#183

Code

Open

obbamna wants to merge 32 commits intoshakes76:topic-recognitionshakes76/PatternAnalysis-2024:topic-recognitionfrom obbamna:topic-recognitionobbamna/PatternAnalysis-2024:topic-recognitionCopy head branch name to clipboard

+803

Commits

Commits on Oct 29, 2024

• initial commit with required files and descriptions

  obbamna

  committed
  ceb154c

  Copy full SHA for ceb154c

Commits on Oct 30, 2024

• added 2d nifti file reader from appendix to dataset.py

  obbamna

  committed
  d865ece

  Copy full SHA for d865ece

• added 2d Unet model to modules.py

  obbamna

  committed
  0c36e72

  Copy full SHA for 0c36e72

• added 2d Unet to modules.py

  obbamna

  committed
  88a0f2a

  Copy full SHA for 88a0f2a

• added dataset class for loading 2d slices

  obbamna

  committed
  9022ce3

  Copy full SHA for 9022ce3

• added training variables

  obbamna

  committed
  ebb80e5

  Copy full SHA for ebb80e5

• added test for unet module

  obbamna

  committed
  7fb82e7

  Copy full SHA for 7fb82e7

• change dataset class to deal with mismatch in size

  obbamna

  committed
  73271e4

  Copy full SHA for 73271e4

• created training loop

  obbamna

  committed
  a600da2

  Copy full SHA for a600da2

• redesign resizing in dataset.py

  obbamna

  committed
  9f30838

  Copy full SHA for 9f30838

• dimension resizing in dataset.py

  obbamna

  committed
  506bb64

  Copy full SHA for 506bb64

• diceloss function added to modules.py

  obbamna

  committed
  6cd03e5

  Copy full SHA for 6cd03e5

• restructure of training with plots of epochs and dice loss

  obbamna

  committed
  43c1eb7

  Copy full SHA for 43c1eb7

• new training restructure

  obbamna

  committed
  251acd4

  Copy full SHA for 251acd4

• changes to Unet to ensure number of channels are conserved when upsampling

  obbamna

  committed
  d165e92

  Copy full SHA for d165e92

• minor changes to dataset.py to include transformations

  obbamna

  committed
  c0e0b7f

  Copy full SHA for c0e0b7f

Commits on Oct 31, 2024

• added more plots and predictions to train.py

  obbamna

  committed
  751df6a

  Copy full SHA for 751df6a

• updated dataset slightly

  obbamna

  committed
  6417225

  Copy full SHA for 6417225

• change to unet structure, max layer is 256

  obbamna

  committed
  e572460

  Copy full SHA for e572460

• changed back conv layers to earlier version

  obbamna

  committed
  5a92ee3

  Copy full SHA for 5a92ee3

• changing hyperparameters and loss function

  obbamna

  committed
  d06e6c5

  Copy full SHA for d06e6c5

• writing README report

  obbamna

  committed
  b21986e

  Copy full SHA for b21986e

• predict script

  obbamna

  committed
  0d03c8a

  Copy full SHA for 0d03c8a

• comments and more pictures in training

  obbamna

  committed
  4c35e4a

  Copy full SHA for 4c35e4a

• finish readme

  obbamna

  committed
  da5b576

  Copy full SHA for da5b576

• minor changes

  obbamna

  committed
  5e463cc

  Copy full SHA for 5e463cc

• minor changes, spelling etc.

  obbamna

  committed
  53f8046

  Copy full SHA for 53f8046

• savedmodels and images for report

  obbamna

  committed
  7d012fc

  Copy full SHA for 7d012fc

• comments

  obbamna

  committed
  6a00afc

  Copy full SHA for 6a00afc

• comments

  obbamna

  committed
  95428b6

  Copy full SHA for 95428b6

• comments

  obbamna

  committed
  3d1359f

  Copy full SHA for 3d1359f

Commits on Nov 14, 2024

• removed model checkpoints

  obbamna

  committed
  4101f98

  Copy full SHA for 4101f98