self-collected VR-MI dataset

The VR-MI (Virtual Reality\u2013based Motor Imagery) dataset is a self-collected EEG dataset designed to explore motor imagery decoding under immersive virtual environments. It contains EEG recordings from 20 right-handed participants performing left- and right-hand motor imagery tasks in VR scenarios using a 32-channel Emotiv Flex 2.0 device at a sampling rate of 256 Hz. Each subject completed 200 trials per condition, with all signals band-limited to 0.5\u201370 Hz and stored in EDF format. The VR-MI dataset provides a valuable benchmark for studying spatio-temporal feature representation, cross-subject generalization, and robustness in EEG-based brain\u2013computer interface research.