THE GOVERNING COUNCIL OF THE UNIVERSITY OF TORONTO
发明人:
NESTOR, ADRIAN RAZVAN,NEMRODOV, DAN
申请号：
CA3044281
公开号：
CA3044281A1
申请日:
2019.05.27
申请国别(地区):
CA
年份:
2019
代理人:
摘要：
There is provided a system and method for generating visual categoryreconstruction fromelectroencephalography (EEG) signals. The method includes: receiving scalp EEGsignals；using a trained pattern classifier, determining pairwise discrimination of theEEG signals, thepattern classifier trained using a training set comprising EEG signalsassociated with a subjectexperiencing different known visual identities； determining discriminabilityestimates of thepairwise discrimination of the EEG signals by constructing a confusabilitymatrix； generating amultidimensional visual representational space； determining visual featuresfor each dimensionof the visual representational space by determining weighted sums of imagestimulusproperties； identifying subspaces determined to be relevant forreconstruction； andreconstructing the visual appearance of a reconstruction target usingestimated coordinates ofthe target and a summed linear combination of the visual features proportionalwith thecoordinates of the target in the visual representational space.
