How did you pre-process the MNIST dataset as it is one channel?

kevinlin311tw / Caffe-DeepBinaryCode

Supervised Semantics-preserving Deep Hashing (TPAMI18)

https://arxiv.org/abs/1507.00101v2

Other

205 stars 83 forks source link

Closed gewen closed 6 years ago

gewen commented 6 years ago

Hi, how did you pre-process the MNIST dataset as it is one channel which is different from the input of the ImageNet pretrained model? Thanks.

kevinlin311tw commented 6 years ago

It is totally fine to duplicate it to other two channels.