Nat Hum Behav | The human motor cortex encodes complex handwriting by transitioning through neural states, each with distinct directional tuning, enabling accurate reconstruction of written characters.
Apr 2, 2025
eLife | A framework is proposed to separate and analyze behaviorally relevant neural signals, revealing that previously overlooked neural responses encode rich information and suggesting that motor behaviors occupy a higher-dimensional space than expected and can be decoded linearly.
Aug 9, 2024
BIBM | This paper proposes SeSA, a semi-supervised domain adaptation framework that improves BCI recalibration by aligning both global and speed-specific neural features, achieving robust decoding across days despite neural nonstationarities.
Jan 1, 2024
NeurIPS | A novel evolutionary ensemble framework, EvoEnsemble, dynamically adapts to changes in neural signals by evolving decoders over time, significantly improving the accuracy and robustness of neural decoding in nonstationary conditions.
Dec 6, 2022
CICAI | An AI-augmented brain-computer interface is developed for a mahjong game, enabling easier and more precise control by allowing users to select from AI-suggested options, thus improving usability and user experience in a real BCI application.
Aug 27, 2022
IEEE Trans on Biomedical Engineering | A dynamic ensemble Bayesian filter (DyEnsemble) is proposed to improve the robustness of online brain-machine interface (BMI) control by adaptively combining multiple neural decoding models to handle neural signal variability, showing significantly better accuracy and stability than conventional methods.
Jun 13, 2022