search

GeometricBCI / Deep-Optimal-Transport-for-Domain-Adaptation-on-SPD-Manifolds

This is the official GitHub repo for "Deep Optimal Transport for Domain Adaptation on SPD Manifolds"
Other
6 stars 0 forks source link

can you provide dataset npy file thank u #1

Open a912289748 opened 1 year ago

GeometricBCI commented 1 year ago

I'm glad you're interested in my work. Your question is not very clear. Are you referring to the .npy file as the dataset or an intermediate variable? If you are referring to the datasets used in the paper, they are all publicly available and can be downloaded. If you are referring to an intermediate variable, you can directly call the code to obtain it. Thank you.

a912289748 commented 1 year ago

thank you, your work is very good for me,i very admire you, the public data do you need process? or direct use

At 2023-08-01 15:27:51, "Ce Ju" @.***> wrote:

I'm glad you're interested in my work. Your question is not very clear. Are you referring to the .npy file as the dataset or an intermediate variable? If you are referring to the datasets used in the paper, they are all publicly available and can be downloaded. If you are referring to an intermediate variable, you can directly call the code to obtain it. Thank you.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

GeometricBCI commented 1 year ago

Thank you for your appreciation. Generally, deep learning (DL) methods greatly simplify the cumbersome signal preprocessing steps, but preprocessing is still necessary. For the three publicly available datasets used in the paper, the preprocessing code can be found in the repository. (These datasets likely underwent some simple processing before being made public.)

Please keep in mind that the main focus of this paper is to propose a novel domain adaptation method rather than solely pursuing classification accuracy. The current simple preprocessing is sufficient for the purpose. However, for your in-house dataset, I recommend using conventional EEG signal processing methods to filter out noise and make necessary corrections before applying all of the DL methods, including the one proposed in this paper.