1 code implementation • 30 Sep 2022 • Sikun Lin, Thomas Sprague, Ambuj K Singh
Instead of training models from scratch to find a latent space shared by the three modalities, we encode fMRI signals into this pre-aligned latent space.
