GaotongYu

我的也是没锚框 Predict全部都没有

Same question, have you solved that?

@seekFire Nice work!Thx for all this!

Same question[1], but question[0] is easy. Before training or testing, you should run the pre_process.py to pre-process the dataset from .mat to .npy.

@seekFire Thx too! Keypoints[:, 2] represents the mean value of three nearest point distances. 都是中国人本来想中文回复的，但是为了国外友人能看懂还是英文回复了 加油 :)

Same problem! Have anyone solved it?

I solved it just add three lines under train()->images, labels = images.cuda(), labels.cuda(): `shape = list(images.shape) images = images.view(shape[0]*shape[1], *shape[2:]) labels = labels.view(-1)` just combine the 1st and the 2nd...