search

Joker316701882 / Additive-Margin-Softmax

This is the implementation of paper <Additive Margin Softmax for Face Verification>
492 stars 149 forks source link

Some problems about align_lfw.py #21

Open Carbord opened 5 years ago

Carbord commented 5 years ago

As is mentioned is the README.md, I should align both the training set and lfw. Is this mean that I should use the align_lfw.py for twice? My setting is like following: Step1 --"python align_lfw.py --input-dir /DataSet/lfwdataset/lfw-deepfunneled --output-dir /home/myname/lfwdataset" Step2 --"python train.py --data_dir /home/myname/lfwdataset --random_flip --learning_rate -1 --learning_rate_schedule_file ./data/learning_rate_AM_softmax.txt --lfw_dir /home/myrname/lfwdataset --keep_probability 0.8 --weight_decay 5e-4" Is there any problems? Hope for your reply soon, thank you~

Joker316701882 commented 5 years ago

@Carbord Well, you can't train on lfw while testing on it. Try to train on MS-celeb-1M or VGGface2.