Bimodal dataset on Inner speech
A bimodal dataset of EEG and fMRI recordings acquired nonsimultaneously during inner-speech production in four healthy, right-handed participants. Participants performed an inner-speech task with eight words from two categories, social and numeric, with 40 trials per word, giving 320 trials per modality per participant. The dataset is intended to support inner speech recognition and brain-computer interfaces for speech prostheses.
- Participants
- 4
- Size
- 9.33 GB
- Version
- vv1.0.0
- Updated
- Sep 30, 2026