Xuan Ma

The paper develops a framework to quantify the mechanics of primate dexterous grasp by measuring contact forces, kinematics, and joint torques using an experimental apparatus and musculoskeletal modeling.

The paper shows that motor cortex neurons encode movement through a consistent muscle-based principle, explaining both reaching and grasping behaviors without distinct control strategies.

This paper introduces NoMAD, a method that stabilizes brain-computer interfaces by aligning latent dynamics, achieving stable decoding without supervised recalibration over long periods.

This paper introduces a discrete direction selection decoder for iBCI that outperforms continuous velocity decoders in human and monkey experiments by improving control accuracy and reducing noise.

This paper develops an unsupervised piecewise linear decoder for multi-task brain-computer interfaces, showing it outperforms global decoders in EMG prediction.

The paper trains an autoregressive Transformer on neural spiking data to decode motor behavior, showing broad utility across tasks but highlighting limitations in scaling.

The paper introduces FALCON, a benchmark suite for evaluating robustness of iBCI decoders using standardized datasets and evaluation metrics.

This paper transfers brain-computer interface decoders from monkeys to humans, showing that motor-related neural representations are consistent across species and reducing the need for extensive data.

This paper improves brain-computer interface decoding accuracy over time using Cycle-GAN and ADAN methods, outperforming linear methods with minimal data requirements.

The paper investigates whether low-dimensional neural manifolds in M1 are present during unconstrained natural movements and finds they are, with similar dimensionality to constrained movements.