Hi MNE community,
I’m Anushka, an Electrical Engineering student at UIUC. I currently do undergraduate research in the Singer Lab, working specifically on EEG-based auditory attention decoding.
For GSoC 2026 I’m proposing to build AAD pipeline tooling inside mne.decoding. An EnvelopeEstimator, a TemporalResponseFunction estimator with sklearn API compatibility, an AuditoryAttentionDecoder class with temporal-autocorrelation-safe cross-validation, and integration of a public AAD dataset via mne.datasets. The goal is to make AAD experiments reproducible and accessible without every lab rebuilding the same boilerplate.
Would genuinely appreciate any mentor feedback on the scope or direction before the deadline today. Happy to share my draft proposal.