Licenses & access¶

LAION-fMRI has two release tracks with different rules:

The fMRI data, derivatives, and metadata are released under CC0 1.0. You can download them anonymously; we just ask you to acknowledge the licence once.
The stimulus images come from third-party web sources and aren’t CC0. We give them to you only after you accept a short Data Use Agreement (DUA): https://laion-fmri.hebartlab.com/terms.

Most of this page is about the DUA flow — the CC0 prompt is a single "I AGREE".

The CC0 prompt¶

The first time you run download(...) against a new data directory, the package shows you the CC0 license and asks you to type "I AGREE":

=== LAION-fMRI Dataset License (CC0 1.0) ===

The brain imaging and participant data in the LAION-fMRI dataset are
released under the Creative Commons Zero (CC0 1.0) Public Domain
Dedication. You are free to copy, modify, distribute, and use the
data for any purpose, including commercial, without asking permission.

Full license text: https://creativecommons.org/publicdomain/zero/1.0/

Type "I AGREE" to accept and continue with the download:

We write a small marker file at {data_dir}/.laion_fmri/license_accepted so you never see the prompt again on that data directory.

If you’d rather review the licence up front:

from laion_fmri.download import accept_license
accept_license()

If you decline, the package raises laion_fmri._errors.LicenseNotAcceptedError and stops.

The DUA flow (stimulus images)¶

The stimulus images come from third-party web sources — we don’t own them. To give them to you we ask you to accept a short Data Use Agreement first. There’s no login, no password, no email verification. You fill in a short form once; the laion_fmri package then signs S3 URLs for you on demand.

How to download¶

From a Python session:

import laion_fmri
laion_fmri.download_stimuli()

…or from the shell:

laion-fmri download-stimuli

The first time, you’ll be prompted in the terminal for your name, institutional email, institution, optional PI/supervisor, and a brief description of your research purpose. The prompt links to both the Terms of Use and Privacy notice; type yes to accept the Terms and acknowledge the Privacy notice, and the download starts (~3 GB).

Running the command again later just re-downloads what’s missing — no form to re-fill.

Reading the images¶

For the full API see Load. The short version:

import laion_fmri
stim = laion_fmri.load_stimuli()
stim.images["shared_12rep_LAION_cluster_1003_i0.jpg"]   # raw JPEG bytes
stim.images.get(0)                                      # PIL.Image

Together with fMRI in one call¶

Pass include_stimuli=True to the regular download() — the fMRI part runs first, the stimuli after:

from laion_fmri.download import download
download(subject="sub-03", include_stimuli=True)

In a browser, without Python¶

If you’d rather not install the package, submit the same form in a browser: https://laion-fmri.hebartlab.com/request. The confirmation page gives you direct download URLs (valid one hour) that you can use with curl or click in your browser.

On a cluster¶

Easiest path: download once on your laptop, then just rsync the data to the cluster:

rsync -av ~/path/to/laion-fmri-data/stimuli/ cluster:~/path/to/laion-fmri-data/stimuli/

On the cluster, load_stimuli() reads the local files without ever calling the access service, and download_stimuli() notices the files already match and returns immediately.

You only need access state on the cluster if you also want to re-download there (e.g. after a dataset release update). In that case set the LAION_FMRI_REQUEST_ID environment variable to the value you can find in ~/.cache/laion-fmri/auth.json on your laptop — but most workflows don’t need this.

If something goes wrong¶

We’ve updated the DUA. The package prints a URL; open it, click I accept, and re-run the command.
The download was interrupted. Just run the command again — it picks up where it left off.
Anything else. Open an issue on GitHub. For stimulus access / takedown questions, see the contact at https://laion-fmri.hebartlab.com/takedown.

Privacy¶

What we keep on the server. Your form submission (name, institutional email, institution, optional supervisor, research purpose) and a record of which files we sent you and when.

What we don’t keep. Your IP address or anything about your browser.

What we never do. Send you email — there’s no email verification, no newsletter, no automated notifications.

What happens to your record over time. After a year with no downloads we automatically anonymise your form submission: name, email, institution, supervisor and research purpose are dropped while a minimal audit row (which files were sent, when) stays for governance.

The full privacy notice — including the data-controller contact — is at https://laion-fmri.hebartlab.com/privacy.