Closed beckhamchen closed 5 years ago
In train_test.py: rgb_std = (1, 1, 1) while rgb_means = (103.94, 116.78, 123.68)
But in train_test_fssd_mobile_pre.py rgb_means = (255 0.485, 255 0.456, 255 0.406) rgb_std = (0.229 255, 0.224 255, 0.255 255)
the two rgt_std are far away. I think rgb_std in train_test.py is not right. Anybody kown ?
The mean and std are same with the pretrained model. For vgg, we aodpt the model from caffe. For mobilenet, we adopt the model which is trained with pytorch.
In train_test.py: rgb_std = (1, 1, 1) while rgb_means = (103.94, 116.78, 123.68)
But in train_test_fssd_mobile_pre.py rgb_means = (255 0.485, 255 0.456, 255 0.406) rgb_std = (0.229 255, 0.224 255, 0.255 255)
the two rgt_std are far away. I think rgb_std in train_test.py is not right. Anybody kown ?