Brain decoding: toward real-time reconstruction of visual perception

Authors: Yohann Benchetrit, Hubert Banville, Jean-Remi King

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our results are threefold: Firstly, our MEG decoder shows a 7X improvement of image-retrieval over classic linear decoders. Second, late brain responses to images are best decoded with DINOv2, a recent foundational image model. Third, image retrievals and generations both suggest that high-level visual features can be decoded from MEG signals, although the same approach applied to 7T f MRI also recovers better low-level features. Overall, these results, while preliminary, provide an important step towards the decoding in real-time of the visual processes continuously unfolding within the human brain.
Researcher Affiliation Collaboration Yohann Benchetrit1, , Hubert Banville1, , Jean-R emi King1,2 1FAIR, Meta, 2 Laboratoire des Syst emes Perceptifs, Ecole Normale Sup erieure, PSL University {ybenchetrit,hubertjb,jeanremi}@meta.com
Pseudocode No The paper describes the architecture and training process in text but does not provide structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any explicit statement or link regarding the release of open-source code for the described methodology.
Open Datasets Yes We test our approach on the THINGS-MEG dataset (Hebart et al., 2023). Four participants (2 female, 2 male; mean age of 23.25 years), underwent 12 MEG sessions during which they were presented with a set of 22,448 unique images selected from the THINGS database (Hebart et al., 2019), covering 1,854 categories.
Dataset Splits Yes We use early stopping on a validation set of 15,800 examples randomly sampled from the original training set, with a patience of 10, and evaluate the performance of the model on a held-out test set (see below). Models are trained on a single Volta GPU with 32 GB of memory. We train each model three times using three different random seeds for the weight initialization of the brain module.
Hardware Specification Yes Models are trained on a single Volta GPU with 32 GB of memory.
Software Dependencies No The paper mentions "Adam optimizer (Kingma & Ba, 2014)" and "Scikit-learn (Pedregosa et al., 2011)" but does not provide specific version numbers for all key software components (e.g., deep learning frameworks or other relevant libraries) used for the experiments.
Experiment Setup Yes Cross-participant models are trained on a set of 63,000 examples using the Adam optimizer (Kingma & Ba, 2014) with default parameters (β1=0.9, β2=0.999), a learning rate of 3 10 4 and a batch size of 128. We use early stopping on a validation set of 15,800 examples randomly sampled from the original training set, with a patience of 10, and evaluate the performance of the model on a held-out test set (see below).