Fighting-JJ

Anyone who has trained the model on dataset syth90k? Could you show the final accuracy ?

10

And the tricks you have tried to prove the accuracy.  the accuracy that i achieve. i just want to know is there any promotion can be done?

Can I build Glow on windows_10 ?

5

Due to the read.me , Glow can only be build on MacOs and Linuxs.

边框回归中的 wh_loss计算部分只对true_box的wh取log运算了，而预测边框的宽高却没有？

2

在yolo_head函数中： box_wh = tf.exp(predictions[..., 2:4]) * anchors_tensor / input_shape[::-1] 在yolo_loss函数中： raw_true_wh = tf.log(tf.where(tf.equal(y_true[index][..., 2:4] / anchors[anchor_mask[index]] * input_shape[::-1], 0), tf.ones_like(y_true[index][..., 2:4]), y_true[index][..., 2:4] / anchors[anchor_mask[index]] * input_shape[::-1])) ... wh_loss =...

The link " collected font dataset " is dead

4

When click the link : https://github.com/azadis/AdobeFontDropper#mc-gan-traintest it turns out to be 404.