Sign in to view source links and access this dataset
Description
38 participants aged 18–35 contributed over 31 hours of EEG recordings during sleep, paired with verbatim dream reports and AI-generated images derived from those reports. The dataset was created by opsecsystems and last updated on April 6, 2026. It is described as the world's first multimodal resource linking these three data types for research at the intersection of neuroscience, psychology, and artificial intelligence.
Use Cases
Train models to decode EEG signals into dream content based on the paired EEG and dream report data.
Develop AI systems that generate images from textual dream descriptions using the provided report-image pairs.
Study the relationship between brain activity patterns and subjective dream experiences using the synchronized EEG and report data.
Benchmark multimodal fusion techniques that integrate time-series physiological data, text, and images.
Strengths
Includes data from 38 healthy participants aged 18–35.
Contains over 31 hours of EEG recordings synchronized with dream reports.
Provides a unique multimodal link between EEG signals, text reports, and AI-generated images.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Row count and file formats are unknown, which may limit suitability assessment.
The dataset's sample size of 38 participants may limit generalizability of findings.
Provenance
Source
opsecsystems via Hugging Face.
Collection Method
Collected from 38 participants during sleep, with subsequent AI processing of dream reports to generate images.
Freshness
Last updated 2026-04-06 22:03:32.
License is unknown; terms of use must be verified before application.