[MRG] Add SPDNet #534
[MRG] Add SPDNet #534
Conversation
|
Ping @rkobler |
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## master #534 +/- ##
==========================================
+ Coverage 84.72% 84.75% +0.02%
==========================================
Files 63 64 +1
Lines 4741 4860 +119
==========================================
+ Hits 4017 4119 +102
- Misses 724 741 +17 |
|
Hi ! |
|
Hi @tgnassou! Great that you started to implement SPD type models in braindecode. I am happy to support and help in the implementation process! @jpaillard it would be definitely interesting to test both approaches. So far, I only have experience with geoopt. |
* moving * Fixing typo and analysing the output * Fixing * removing env * Updating whats new * dummy commment * Cleaning
* Set signal-related parameters in check_data * Merge tests for EEGClassifier and EEGRegressor * Fix coquille * Update test to use module mixin * Rename clf to eegneuralnet in test * Add preds fixture * Update eegneuralnet.py and subclasses * Update test_eegneuralnet.py * restored previous tests behavior where nn layer is ignored and always returns mocked values * Update whats_new.rst * Use two different mock modules for test * Rename mock modules * Use set_params instead of vars * Deprecate passing an initialized module and skip setting signal args in that case * Remove unnecessary f-strings * Try fix python 3.8 * Fix case with non torch dataset * Try fix python 3.8 * Add docstrings for fit and partial_fit (already including braindecode#529) * Test initialized module * Fix Flake8 * Add email * Remove deprecation of initialized module --------- Co-authored-by: Maciej Sliwowski <macieksliwowski@gmail.com> Co-authored-by: Bru <a.bruno@aluno.ufabc.edu.br>
# Conflicts: # braindecode/models/functions.py # docs/whats_new.rst # setup.py # test/unit_tests/models/test_models.py
|
Things exploded when I updated the branch. I will re-start in a fresh branch/PR @tgnassou |
Hi, I started implementing SPDNet in Braindecode, as discussed in the issue #453.
It is just a draft, and the model is not working. I added an example if someone wants to try it. I still need the pooling and batch normalization layers to have the same SPDNet as Huang et al.. I will also add the Riemannian batch norm of Kobler et al. and Brooks et al..
I am new in this area of SPDNet. If anyone has any insights or suggestions, please let me know.