search

KamitaniLab / GenericObjectDecoding

Demo code for Horikawa and Kamitani (2017) Generic decoding of seen and imagined objects using hierarchical visual features. Nat Commun https://www.nature.com/articles/ncomms15037.
150 stars 46 forks source link

About ImageFeature file #13

Closed mina98 closed 4 years ago

mina98 commented 4 years ago

i want to ask about more details for image feature file , how it's shape is (16234,) but the stimulus photos are only 1250 image

mitsuaki commented 4 years ago

In addition to the features for 1250 images, we included category-averaged features for 15,372 categories in ImageNet (See "Category-average feature decoding" subsection in Horikawa and Kamitani (2017), https://www.nature.com/articles/ncomms15037 ). So, totally we have 16622 image feature vectors in the image feature file (ImageFeature.h5). If not, the file may be broken.

mina98 commented 4 years ago

okay thanks, can i get the images, stimulus and visual images? because some of this photo deleted on imagenet @mitsuaki

mina98 commented 4 years ago

Thanks for helping