aimerykong / Low-Rank-Bilinear-Pooling

CVPR2017 - an ultra-compact bilinear model for fine-grained classification
https://www.cs.cmu.edu/~shuk/lr_bilinear.html
149 stars 36 forks source link

LRBL accuracy issues #9

Open dadongwang opened 5 years ago

dadongwang commented 5 years ago

I see from the demo 4 that the final accuracy can't reach 84.21% as claimed in the paper? But the caffe-20160312 reports 84.98% for the compact bilinear model already. Could you please explain the difference?