ncclabsustech/deepseparator

Deep learning model for EEG artifact removal

deepseparator's Issues

From the given code file, it seems that the data set is not divided into training set, test set and verification set, and the optimal model is directly selected on the verification set, which has not been verified by the test set.

Signal to Noise Ratio

Hi, I was wondering what SNR results are you showing on table 1 in the article? Thanks.

The data set I use is the data set disclosed in the EEGdenoisednet article,

I would like to ask whether the parameters in the deep separator project are the corresponding parameters of the performance reported in the arXiv article. The results of removing EMG noise using this model are inconsistent with those described in the article.

Incomplete code

Hi,
I was able to run the code provided here and could see one test output getting denoised. But, I wanted to reproduce the results (plots, comparison table) as described in the arxiv paper. It seems like some code and data is not available right now.

Can you please guide me how can I achieve that so that I can extend your work.

Thankyou