Update main - #70

Merged

CfM47 merged 208 commits into

mainfrom

Jan 10, 2026

Commits on Nov 15, 2025

Commits on Nov 16, 2025

feat: add basic scaffolding, simple cnn approach for classification
dhc667
committed
b5968c3
docs(REAMDE): add data setup instructions
dhc667
committed
8812dc8
feat: add data exploration notebook
dhc667
committed
345a121
fix(tests): add pytest fix
dhc667
committed
e2054e7
refactor(config): separate configuration loading from approaches
dhc667
committed
9a5bf58
style(cnn_v1): rename config file for clarity
dhc667
committed
86bb98d

Commits on Nov 21, 2025

Merge pull request #26 from CfM47/feat/add-scaffolding--cnn-classifier
dhc667
authored
ec3334d

Commits on Dec 19, 2025

Commits on Dec 20, 2025

feat: Implement ViT and Swin models, and update project configuration.
fprestamo
committed
a0f3c0e
feat: Add AutoML Orchestration.
fprestamo
committed
4cd650f

Commits on Dec 26, 2025

refactor(auto-ml): Fix typing, linting errors, make verifications tests
dhc667
committed
430cc4b
feat: Update test_dataset_loader and test_pipeline

fprestamo
authored and
dhc667
committed
c1e304c
fix: delete structure
dhc667
committed
c8e64cc
fix: delete old approach in favor of auto ml
dhc667
committed
64d1b6b
Merge pull request #28 from CfM47/feat/auto-ml
dhc667
authored
c122918
feat: Update auto_ml components, main script, and tests.
fprestamo
committed
2a243fa
feat: Update model test files.
fprestamo
committed
29079b6
fix(linting): resolve ruff linting errors
theleywin
committed
8179211
Merge pull request #29 from CfM47/feat/add-automl-evaluator-node
theleywin
authored
1d5cb7c

Commits on Dec 28, 2025

feat: Add QuadtreeSegmentationModel and ClassificationModelInterface for image segmentation
CfM47
committed
07e3968
Add autoencoder-based mask evaluator and tests

committed
4b4888d
fix(linting): resolve ruff linting errors
theleywin
committed
4a6a755
test(quadtree): Add tests for quadtree segmentation
CfM47
committed
83bd91b
fix(types): resolve mypy typing errors
theleywin
committed
3388dcf
Merge pull request #32 from CfM47/feat/add-mask-coherence-evaluator
theleywin
authored
f36612a
Merge branch 'dev' into feat/quadtree
theleywin
authored
aa0d44a
Merge pull request #30 from CfM47/feat/quadtree
theleywin
authored
2ed9535
refactor: split implementations.py into modular structure
theleywin
committed
3a9bd2b
refactor: change implementations/models -> implementations/segmentators
theleywin
committed
aa63bed
Merge pull request #33 from CfM47/refactor/implementations
fprestamo
authored
edfb258

Commits on Dec 29, 2025

feat: add CNN classifier and validator notebook
dhc667
committed
ddeba71
fix: appease the makefile
dhc667
committed
39f1593
feat: create classification model interface implementation with CNN
dhc667
committed
df9fb85
test: add quad tree + CNN Classifier integration test
dhc667
committed
b0fd91c
Merge pull request #31 from CfM47/feat/cnn-classifier-and-validator
fprestamo
authored
8f4122c
refactor: rename segmentator models implementations file names, appease makefile
dhc667
committed
4a21769
Merge pull request #34 from CfM47/refactor/model-names--appease-makefile
dhc667
authored
034cd05
feat: implement comprehensive augmentation system with SEM-specific transforms

theleywin
and
DiegoViera1511
committed
8c59ab2
feat: add train method for classifier model interface

dhc667
and
LoLProM
committed
4b9ba3e
fix: make tests deterministic
dhc667
committed
cfe32dc
doc: augmentators docs
DiegoViera1511
committed
e811f5f
doc: add an USAGE_GUIDE and create a modular structure for augmentators doc
theleywin
committed
9d4b370
Merge branch 'dev' into feat/augmentators
theleywin
authored
c50ab06
fix: mypy issues
theleywin
committed
f39993c
Merge branch 'feat/augmentators' of https://github.com/CfM47/ML-Project into feat/augmentators
theleywin
committed
a521e25

Commits on Dec 30, 2025

Merge pull request #36 from CfM47/feat/augmentators
fprestamo
authored
7a3a67c
feat: add train method for classifier model interface

authored and
dhc-selector
committed
444a369
fix: make tests deterministic

dhc667
authored and
dhc-selector
committed
eca1171
feat(datasets): add function to load classification dataset
CfM47
committed
26393c7
feat(quadtree): implement train function in quadtree segmentation model
CfM47
committed
76b15d9
fix: rebase conflicts
dhc667
committed
aa73682
Merge branch 'feat/train-method-for-classifier-interface' of https://github.com/CfM47/ML-Project into feat/train-method-for-classifier-interface
dhc667
committed
b702d40
fix(quadtre): train classifier argument of invalid type
CfM47
committed
461757a
fix: put quadtree segmentator classifier training dataset path in init
dhc667
committed
58203b1
Merge pull request #35 from CfM47/feat/train-method-for-classifier-interface
dhc667
authored
4aa179b
feat: adding vit classifier and created unit tests for ViT classifier and segmentation model to ensure functionality and correctness
LoLProM
committed
659a5f3
fix: fixing linting errors
LoLProM
committed
bdae5b9

Commits on Dec 31, 2025

feat: enhance ViT model with training functionality and adaptive batch sizing
LoLProM
committed
bd05456
fix: typing issue
dhc667
committed
585b08e
Delete unnecesary cast
fprestamo
authored
ac6c838
Remove unused import 'cast' from interfaces.py
fprestamo
authored
c167e9a
Merge pull request #37 from CfM47/feat/vit-classifier
fprestamo
authored
842466d
fix: Save metrics distribution instead of jsut one value.
fprestamo
committed
cedc8d4
feat: implement 18 segmentation evaluators with per-class and macro-average metrics
Sekai02
committed
b121fb2
feat: Implement AutoML caching and autosaver functionality with tests
LoLProM
committed
e99710b
fix: fixing returning any error
LoLProM
committed
e9446c3
refactor: split segmentation.py into separate metric files
Sekai02
committed
dc083ed
Merge pull request #38 from CfM47/feat/add-automl-autosave
fprestamo
authored
2c8d7ef
fix: make automl folder visible.
fprestamo
committed
852c095
Update AutoML result keys and experiment parameters
fprestamo
committed
efdb580

Commits on Jan 1, 2026

fix: ruff issues
pablo-bego
committed
c62844f
fix: disable test for IoUClass0Evaluator
pablo-bego
committed
279c8b8
Merge branch 'dev' into feat/segmentation-models-evaluators
theleywin
authored
a82620b
Fix test
fprestamo
authored
286b022
Merge pull request #39 from CfM47/feat/segmentation-models-evaluators
theleywin
authored
ddc093b
Merge branch 'dev' into fix/automl-autosave
fprestamo
authored
39be605
Fix ruff
fprestamo
authored
d393fc3
Merge pull request #41 from CfM47/fix/automl-autosave
theleywin
authored
8c75447
test(quadtree): add tests for hyperparameter tuning
CfM47
committed
86838a9
Add simulated annealing for quadtree hyperparameter tuning

fprestamo
and
CfM47
committed
afb59e1
Refactor quadtree hyperparameter tuning and tests

fprestamo
and
CfM47
committed
b99e392
Merge pull request #42 from CfM47/feat/quadtree-tunning
fprestamo
authored
acbf74d
feat: add auto ML setup scaffolding

dhc667
authored and
fprestamo
committed
3088387
Implement evaluator and model node setup functions
fprestamo
committed
334d7c6
feat: initialize evaluator node with dataset
fprestamo
committed
702112a
feat(model-setup): instantiadted quadtree nodes per classifier
CfM47
committed
9e51122
fix(model-setup): fix typo in metric to optimize quadtree
CfM47
committed
1126784
Merge branch 'feat/add-auto-ml-setup' of https://github.com/CfM47/ML-Project into feat/add-auto-ml-setup
dhc667
committed
f7348d3

Commits on Jan 2, 2026

fix: make quad tree model train CNN with 50 epochs
dhc667
committed
c14b7cb
feat: Add training and val history and fix an error during trains and validations

fprestamo
authored and
dhc667
committed
0c0e72e
Merge pull request #40 from CfM47/feat/add-auto-ml-setup
fprestamo
authored
58d881f
Merge branch 'dev' into feat/add-training-val-history
fprestamo
authored
e9fc0a4
Merge pull request #43 from CfM47/feat/add-training-val-history
fprestamo
authored
5b4b348
fix: dataset path injection
dhc667
committed
5212947
fix: inject auto ml cache dir
dhc667
committed
4d0073f
fix: add logging for classification models
dhc667
committed
da2b27a
fix: standardize logging for segmentator models
dhc667
committed
901566f
fix: typing issues
dhc667
committed
f961d1c
fix: typo
dhc667
committed
39e9103
Merge pull request #45 from CfM47/feat/add-auto-ml-setup
dhc667
authored
4e514c7
feat: get augmentators nodes function

DiegoViera1511
and
theleywin
committed
2a5983b
Add model hyperparameter customization for ViT and Swin
fprestamo
committed
bbf89fb
Merge pull request #46 from CfM47/fix/vit-and-swim-hyperparameters
dhc667
authored
4975e90
feat: add kaggle notebook
dhc667
committed
ef1d921
Merge pull request #47 from CfM47/add-kaggle-notebook
fprestamo
authored
b195dd3
fix: main path fix
dhc667
committed
bf2f033
Merge pull request #49 from CfM47/fix/paths
fprestamo
authored
eda1a26
fix: missing path type hint
dhc667
committed
6f68c45
update: augmentators functions

theleywin
and
DiegoViera1511
committed
9fc1c1e
fix: ruff linting on baseline

theleywin
and
DiegoViera1511
committed
e37c67a
Merge branch 'dev' into feat/add-auto-ml-setup
theleywin
authored
bbbb06c
feat: Add initial image dataset, sampling script, and AutoML cache.
fprestamo
committed
8ebdca7
fix batch size error
fprestamo
committed
47b7540
fix: epochs error
dhc667
committed
4d4c419
Merge pull request #51 from CfM47/feat/add-auto-ml-setup
fprestamo
authored
95f77be
Merge branch 'dev' into fix/paths
fprestamo
authored
145da49
Merge pull request #50 from CfM47/fix/paths
fprestamo
authored
e519c86
feat: add auto ml node filtering
dhc667
committed
7c7fd82
Merge pull request #52 from CfM47/feat/filter-automl-nodes
fprestamo
authored
ceb2c59
refactor: replace baseline node with identity augmentator in get_augmentator_nodes
fprestamo
committed
f5422ce
Merge pull request #53 from CfM47/feat/fix-baseline-data-augmentator
dhc667
authored
6b54488
fix: negative strides on fliplr function

theleywin
and
DiegoViera1511
committed
a5f465c
Merge pull request #54 from CfM47/fix/flip-augmentators
dhc667
authored
9316db8
feat(cnn): add configurable num_blocks parameter for smaller input sizes
dhc667
committed
1dcb552
feat(quadtree): add detailed logging for training and evaluation processes
fprestamo
committed
0bc2e13
feat(models): update ViT and Swin model configurations for improved performance
fprestamo
committed
453d67a
Merge pull request #55 from CfM47/fix/cnn-smaller-inputs
fprestamo
authored
deb6300
Merge branch 'dev' into feat/add-biger-models
fprestamo
authored
4b4f851
Merge pull request #56 from CfM47/feat/add-biger-models
dhc667
authored
1a1023e
feat(models): addded swin classifier model
CfM47
committed
7175c0e
feat: add sliding window segmentation model with majority voting
Sekai02
committed
9a469ad
fix: resolve line length errors in sliding window tests
Sekai02
committed
cc8e39c
fix: resolve mypy error in sliding window segmentation
Sekai02
committed
64073ce
fix: correct assertions in sliding window tests
Sekai02
committed
8696772

Commits on Jan 3, 2026

refactor: make swin model names consistent
dhc667
committed
8b05022
feat: add confidence-weighted voting to sliding window segmentation
Sekai02
committed
ef6d71e
fix: add trailing comma in sliding window segmentation
Sekai02
committed
b3336c1
fix: adjust confidence weighted test for boundary effects
Sekai02
committed
575a2f9
docs: update aggregation method description and enhance metrics documentation
Sekai02
committed
7e9e17d
Merge pull request #59 from CfM47/feat/sliding-window-segmentation
Sekai02
authored
b247abb
Add model training results and execution times cache

committed
d1a5e6c
Add script to recalculate macro metrics in results cache

committed
fac2ae1
feat: add evaluation metrics and training data for new models in results cache
fprestamo
committed
9bce1db
feat: update evaluation metrics with improved values in results cache
fprestamo
committed
611d218
feat: implement F1 mean calculation and save results to JSON
fprestamo
committed
e6098bf
feat: add final model training setup
dhc667
committed
b23ca35
feat: add patience-based training for swin model
dhc667
committed
533b300
feat: add validation dataset split for final training
dhc667
committed
7938f6d
feat: add kaggle notebooks
dhc667
committed
a667a23
fix: remove branch from kaggle notebooks for mergeability
dhc667
committed
d06edb0
docs: update readme
dhc667
committed
5d8dcad
fix: linting
dhc667
committed
e0c79d7
fix: training config defaults
dhc667
committed
ddf97b8
feat: improve images
dhc667
committed
11cc0ca

Commits on Jan 4, 2026

fix: track best val model even if patience is not set
dhc667
committed
f588d33
refactor: move result processing scripts to results/scripts
dhc667
committed
51e8025
Merge pull request #64 from CfM47/refactor/add-results
dhc667
authored
61c1b55
Merge pull request #58 from CfM47/feat/swin-classifier
dhc667
authored
9af1525
Merge branch 'dev' into feat/add-final-model-training-setup
theleywin
authored
c7b41d4
fix: typing issue
dhc667
committed
51d5a39
Merge branch 'dev' into feat/add-results
theleywin
authored
d635438
Merge pull request #62 from CfM47/feat/add-final-model-training-setup
theleywin
authored
c3c84c6
Merge pull request #63 from CfM47/feat/add-results
theleywin
authored
02a18f4

Commits on Jan 6, 2026

feat: add training history graph
dhc667
committed
9448574

Commits on Jan 8, 2026

docs: add state of the art
CfM47
committed
89e5799
Merge pull request #67 from CfM47/feat/add-training-history-graph
CfM47
authored
abb35aa
Merge pull request #66 from CfM47/feat/soa
CfM47
authored
6b6c8eb
feat(report): added abstract and introduction to report
CfM47
committed
ced9d84
feat(report): add quadtree section
CfM47
committed
f2da2db
feat(report): add data augmentation section to report
CfM47
committed
73c4526
feat(report): add vit and swin segmentation sections to report
CfM47
committed
efccba7
feat(report): add some cohesive paragraphs
CfM47
committed
acb830d
feat(report): add vit classifier section
CfM47
committed
3bc261a
feat(report): add evaluation and metrics section
CfM47
committed
f6d41e1
feat(report): add automl metodology section
CfM47
committed
30d071a
fix(report): improved vit segmentation subsection
CfM47
committed
4397601
fix(report): improved swin semgentation subsection
CfM47
committed
492c65c
fix(report): fix quadtree section
CfM47
committed
4331888
fix(report): fix vit classifier section
CfM47
committed
50b13a8

Commits on Jan 9, 2026

feat(report): add automl results
CfM47
committed
36d8d05
feat: add train loss and val loss plots for models comparitions in training per epoch
fprestamo
committed
7a968a4
Merge branch 'feat/report' of https://github.com/CfM47/ML-Project into feat/report
fprestamo
committed
bd3cc21
Add average fold plots to training and comparison visualizations
fprestamo
committed
86edbca
feat(report): add dataset description
CfM47
committed
6dc49b5
feat: add analisis de Curvas de Aprendizaje
fprestamo
committed
7e02888
feat:add loss vs data size analysis
fprestamo
committed
447cfae
feat: add more analysis
fprestamo
committed
b275adf
feat: update main.tex
fprestamo
committed
3460f19
feat: update report
fprestamo
committed
a764b78
feat(results): add more metrics to results
CfM47
committed
877bf34
feat:update report
fprestamo
committed
9aca74f
Merge branch 'feat/report' of https://github.com/CfM47/ML-Project into feat/report
fprestamo
committed
68c40e4
feat: update report
fprestamo
committed
aef13ad
feat:update report
fprestamo
committed
99fb876
feat(report): update
CfM47
committed
c30544a
Merge pull request #61 from CfM47/feat/report
CfM47
authored
af20284
feat(notebooks): update data exploration with training dataset
CfM47
committed
387fcdf
feat(notebooks): properly name labels
CfM47
committed
7f990cc

Commits on Jan 10, 2026