Siamese network for melanoma classification - 48515739#292
Open
MelMaillot wants to merge 15 commits intoshakes76:topic-recognitionfrom
Open
Siamese network for melanoma classification - 48515739#292MelMaillot wants to merge 15 commits intoshakes76:topic-recognitionfrom
MelMaillot wants to merge 15 commits intoshakes76:topic-recognitionfrom
Conversation
…ase for my solution. Prompts and complete answers from Google's Gemini are included for completeness and transparency
…ng in current state (debugging and running happens on another machine, where repo needs to be cloned to be accessed). The code mostly mixes the Gemini outputs, with additional self-coded parts. Proper data storage and loading still mostly missing
…ining to normal triplet loss (models didn't learn) - more complex embedding network (now a modified ResNet50) - simplified the classifier head - oversample training data but sample the training set for faster training
…etting, results are mildly better
…evious commit, previous commit message applies to this commit as well
…in.py + cleaned up imports in the other files
…ts, along with a few other details
MelMaillot
changed the title
Collaborator
This is an initial inspection, no action is required at this pointRecognition Problem : total : 19.5
Good design: design is fairly strong(1) Overall Great work, thanks Melissa ! |
Collaborator
Marking
Marked as per the due date and changes after which aren't necessarily allowed to contribute to grade for fairness. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request implements a Siamese network for the classification of the ISIC 2020 Kaggle Challenge data set, a dataset of benign and malignant skin lesions.
Th code implements this functionality in python, using the Pytorch library. It introduces four files:
dataset.pyfor data management (splitting augmentation and retrieval)modules.pyfor network architecture and loss functions (ResNet50 with classifier head, triplet loss)predict.pyfor model evaluation (Accuracy, ROC, AUPRC, t-SNE visualisation)train.pyfor training functions and main functionality (main function trains and tests a model according to hyperparameters specified in the file)Please refer to the README file for more details on the functionality as well as a discussion on model results.
Thank you
Melissa Maillot - 48515739