search

amirhfarzaneh / dacl

Deep Attentive Center Loss
61 stars 13 forks source link

question about the pretrain models #6

Open elk-april opened 3 years ago

elk-april commented 3 years ago

Hi, after reading the paper, I found there are no ablation experiments between pre-trained by imagenet and ms-celeb. Could you please tell me the gap between these two methods? Looking forward to your reply.