Skip to content

Latest commit

 

History

History
28 lines (16 loc) · 1.23 KB

README.md

File metadata and controls

28 lines (16 loc) · 1.23 KB

Reconstructing Visual Scene Images from fMRI

Dataset

The dataset has been procured from

ILSVRC2012

The ILSVRC2012 dataset is a subset of the large hand-labeled ImageNet dataset. The validation and test data consist of 150,000 images, and the training data contains 1.2 million images. All the images are hand labeled with the presence or absence of 1000 object categories which do not overlap with each other.

fMRI on ImageNet

The fMRI data were recorded while subjects were viewing object images (image presentation experiment) or were imagining object images (imagery experiment). In the training image session, a total of 1,200 images from 150 object categories were each presented only once. In the test image session, a total of 50 images from 50 object categories were presented 35 times each. All images were from ImageNet (Fall 2011 release). In this paper, we use the fMRI data of the visual cortex (VC) region in the image presentation experiment. The number of voxels in the VC region is 4466. The number of fMRI samples available is 2700.

Model Pipeline

  • Extraction and Analysis
  • Preparation
  • Model Training
  • Evaluation & Validation
  • Serving

Miscallaneous

Authors