Generic Object Decoding: Preprocessed fMRI and Image Features for Visual Perception

Kamitani Lab provides preprocessed fMRI data and image features from a study on generic decoding of seen and imagined objects. The dataset includes data from 5 subjects, processed with both SPM5 and fMRIPrep pipelines, and is 22.2 GB in size. It was last updated on June 3, 2026.

Use Cases

• Training brain decoding models based on preprocessed fMRI signals from image presentation experiments.
• Comparing neural representations of seen versus imagined objects based on the separate imagery experiment data.
• Benchmarking fMRI preprocessing pipelines using the provided SPM5 and fMRIPrep processed versions.
• Developing visual feature reconstruction algorithms based on the hierarchical image features linked to the fMRI data.

Strengths

• Data from 5 human subjects is included, providing a basis for multi-subject analysis.
• Two distinct preprocessing pipelines (SPM5 and fMRIPrep) are provided, enabling methodological comparison.
• Dataset size is 22.2 GB, indicating a substantial volume of neuroimaging data.
• Data is provided in a standardized format (BData) with associated anatomical images and space templates.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect temporal or methodological bias inherent to the original experimental study.

Provenance

Source

Kamitani Lab, based on data from Horikawa & Kamitani (2017) Nat Commun.

Collection Method

fMRI data collected during image presentation and imagery experiments, preprocessed with SPM5 and fMRIPrep.

Freshness

Last updated 2026-06-03 08:33:32; freshness should be verified.