Abstract Speech brain-computer interfaces (BCIs) enable communication with the external world by decoding neural signals. However, language function as a higher-order brain function, the neural mechanisms underlying speech production remain incompletely understood. Currently most existing Chinese EEG datasets use sentences as stimuli, overlooking that Pinyin constitutes the phonetic foundation of Chinese characters, which limits research on decoding individual Chinese character components. Moreover, most datasets employ only one speech production paradigm, preventing exploration of the brain’s diverse speech production modes. This study aims to construct the 3M-CPSEED Chinese Pinyin dataset for exploring neural activity during three distinct speech modes (overt speech, silently articulated speech, imagined speech)of syllables from distinct articulatory positions. The dataset comprises EEG recordings from 20 participants completing four experimental blocks within one day, yielding 1,800 validated trials. 3M-CPSEED holds significant implications for speech neurophysiology research, not only facilitating exploration of neural activity differences across pinyin articulations but also enabling robust transfer learning studies for other alphabetic languages.
Ma et al. (Mon,) studied this question.