dimension of output feature.

seathiefwang / MGN-pytorch

Reproduction of paper: Learning Discriminative Features with Multiple Granularities for Person Re-Identification

285 stars 108 forks source link

Closed KrystalCWT closed 4 years ago

KrystalCWT commented 4 years ago

I check the output of extract_feature and found the dimension is 2048, why? in paper, that's 256