Riemannian Geometry-Preserving Variational Autoencoder for MI-BCI Data Augmentation

Viktorija Poļaka · 2026

Read Riemannian Geometry-Preserving Variational Autoencoder for MI-BCI Data Augmentation by Viktorija Poļaka (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-10563.txt.

Description

Viktorija Poļaka builds a model that makes new brain signal data while keeping its geometry intact, for motor imagery brain computer use. The data are covariance matrices from scalp readings of imagined movement, and those matrices must keep a strict shape. Her autoencoder uses geometric maps and a combined loss for distance, reconstruction, and diversity. It made valid matrices with a latent space that does not depend on the subject, and the new data proved useful in tests.

Catalog record

CREATORViktorija Poļaka
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24459
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.10563

Browse the full library · About the Tower of Babel