MindEye2: Shared-Subject Models Enable fMRI-To-Image With 1 Hour of Data

TL;DR

MindEye2 introduces a shared-subject model that enables high-quality fMRI-to-image reconstructions with only one hour of training data by leveraging cross-subject pretraining and innovative alignment techniques.

Contribution

The paper presents a novel functional alignment and shared latent space approach that significantly reduces training data requirements for fMRI-to-image reconstruction.

Findings

Achieves state-of-the-art reconstruction metrics with minimal data

Demonstrates effective cross-subject generalization

Enables accurate perception reconstruction from a single MRI session

Abstract

Reconstructions of visual perception from brain activity have improved tremendously, but the practical utility of such methods has been limited. This is because such models are trained independently per subject where each subject requires dozens of hours of expensive fMRI training data to attain high-quality results. The present work showcases high-quality reconstructions using only 1 hour of fMRI training data. We pretrain our model across 7 subjects and then fine-tune on minimal data from a new subject. Our novel functional alignment procedure linearly maps all brain data to a shared-subject latent space, followed by a shared non-linear mapping to CLIP image space. We then map from CLIP space to pixel space by fine-tuning Stable Diffusion XL to accept CLIP latents as inputs instead of text. This approach improves out-of-subject generalization with limited training data and also attains state-of-the-art image retrieval and reconstruction metrics compared to single-subject approaches. MindEye2 demonstrates how accurate reconstructions of perception are possible from a single visit to the MRI facility. All code is available on GitHub.