I joined the lab for my bachelor thesis (B.Sc. Bioinformatics) in 11/2019. In my thesis (Reconstructing visual stimuli from brain activity data using deep neural networks) I used convolutional and recurrent neural networks to reconstruct certain features such as image depth or shapes of visual stimuli from brain activity data (both EEG and fMRI).
Since 05/2021 I am continuing that work mostly on EEG data. Building on the work I have done for my thesis, we have developed an optimized set of computer-generated visual stimuli. The use of artificially created stimuli allows for well controlled training and testing datasets and provides perfect ground truth data, with which the networks can be trained.