experiencor / keras-yolo2

Easy training on custom dataset. Various backends (MobileNet and SqueezeNet) supported. A YOLO demo to detect raccoon run entirely in brower is accessible at https://git.io/vF7vI (not on Windows).

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

issue with model layer lambda,

speeding-motor opened this issue · comments

in fronted 68 line:

output = Lambda(lambda args: args[0])([output, self.true_boxes])
self.model = Model([input_image, self.true_boxes], output)

what that mean? I m complete confuse with this part,i m great appericate you answer me

lambda layer is only for training, for inference you can remove it,
#165 (comment)