Skip to content

Dataset · Fudan-fMRI

Multimodal brain activity recordings during video watching

Fudan-fMRI/CineBrain

Large fMRI, EEG, and ECG signal dataset for neuro-decoding research and multimodal learning.

What's inside

It collects functional MRI (fMRI), electroencephalography (EEG), and electrocardiogram (ECG) recordings of participants watching episodes of the sitcom The Big Bang Theory. The data volume ranges from 100K to 1M entries, and the material language is English. The dataset was presented at CVPR 2026 and is designed for studying audiovisual narrative processing.

What you can do with it

  • Training video and image reconstruction models from brain activity signals
  • Researching modality transfer and multimodal learning
  • Analyzing human brain reactions to natural audiovisual narratives
Tasks
Text → video, Image → video
Languages
English
Size
100K–1M rows
Downloads this month
1.9K
Likes
13

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("Fudan-fMRI/CineBrain")