Multimodal Image and Visually Evoked EEG Dataset for Multiclass Visual Content Classification

The dataset comprises EEG recordings from 20 healthy participants (10 female, 10 male, aged 20–35) collected using a Mitsar-202-24 system with a 250 Hz sampling rate. Signals were captured from six electrodes (Fp2, F3, F4, Fz, Cz, Pz) following the 10–20 international system, focusing on frontal, central, and parietal regions critical for visual processing. Participants were exposed to 120 images across four categories (animals, food, office supplies, vehicles), with each image displayed for 3 seconds followed by a 1.5-second inter-stimulus interval. Data preprocessing included bandpass filtering, CSP, and LMS filtering to enhance signal quality. The dataset supports multiclass visual content classification using deep learning architectures. Access is restricted due to ethical constraints but available upon reasonable request.