The Algonauts Project 2019 involved both training and test datasets. All training data was included within the original dataset and development kit released to participants on April 1, 2019. This set also included the 78 test set images (but no fMRI or MEG data). The challenge test data (twinset1) was not available to participants during the competition and was used to rank submissions. Similarly, the Hidden Test Set (twinset2) was not available to participants and was used to rank a single submission after the main challenge had ended to assess the degree of overfitting/generalizability of the submitted models. Here, we release all datasets at once.

The data belong to a study explained on this webpage (Mohsenzadeh et al., 2019) which used fMRI and MEG to record data while 15 participants viewed a set of 156 natural images. These images can be subdivided into five categories (faces, bodies, animals, objects, scenes) or two twinsets of 78 images each. Images in twinsets 1 and 2 are each composed of unique images that share the same exemplar (e.g. both twinsets each have a unique image of a giraffe). Beta maps from the fMRI data and the MEG epoched data per subject are available to download from that website.

Download the original training dataset, test images, and development kit released on April 1, 2019 to participants of the Algonauts Project Challenge 2019 here.
Below is a list of what is included:

• Challenge Training Set A: 92 images (.jpg format) + fMRI/MEG human brain data RDMs from 2 regions/time intervals (EVC and IT/early and late) in response to viewing images from this set (.mat format).
• Challenge Training Set B: 118 images (.jpg format) + fMRI/MEG human brain data RDMs from 2 regions/time intervals (EVC and IT/early and late) in response to viewing images from this set (.mat format).
• Development Kit: Sample code for generating activations from AlexNet/VGG/ResNet, computing RDMs from these networks, and evaluating model RDMs.
• Challenge Test Set Images: 78 images (.jpg format).

Download all data from the Challenge Test Set and the Hidden Test Set used in the Algonauts Project Challenge 2019 at this link.
Below is a list of what is included:

• Challenge Test Set RDMs: fMRI/MEG human brain data RDMs from 2 regions/time intervals (EVC and IT/early and late) in response to viewing images from this set (.mat format).
• Hidden Test Set RDMs: fMRI/MEG human brain data RDMs from 2 regions/time intervals (EVC and IT/early and late) in response to viewing images from this set (.mat format).
• Challenge and Hidden Test Set Images: 78+78=156 images (.jpg format).

Citation

If you use any of the above data, please cite the Algonauts Project 2019:

Radoslaw Martin Cichy, Gemma Roig, Alex Andonian, Kshitij Dwivedi, Benjamin Lahner, Alex Lascelles, Yalda Mohsenzadeh, Kandan Ramakrishnan, and Aude Oliva. (2019). The Algonauts Project: A Platform for Communication between the Sciences of Biological and Artificial Intelligence. arXiv, arXiv:1905.05675. PDF | arXiv cs.CV

If you use the training data specifically, please also cite:

Radoslaw M. Cichy, Dimitrios Pantazis and Aude Oliva. (2016). Similarity-Based Fusion of MEG and fMRI Reveals Spatio-Temporal Dynamics in Human Cortex During Visual Object Recognition. Cerebral Cortex, 26 (8): 3563-3579. PDF | Cerebral Cortex

If you use the test data specifically, please also cite:

Yalda Mohsenzadeh*, Caitlin Mullin*, Benjamin Lahner, Radoslaw Martin Cichy, and Aude Oliva. (2019). Reliability and Generalizability of Similarity-Based Fusion of MEG and fMRI Data in Human Ventral and Dorsal Visual Streams. Vision, 3(1), 8. PDF | Vision