songhengyang/face_landmark_factory

Issues

• correct_pad

  #13 opened 4 years ago by jeranio
  4

• train color image

  #21 opened 3 years ago by soans1994
  0

• Bug in augment script

  #17 opened 4 years ago by Source82
  2

• data augmentation

  #20 opened 4 years ago by soans1994
  0

• loss function error

  #19 opened 4 years ago by soans1994
  0

• difference between landmark tensorflow implementation and openpose/tfpose for human pose estimation

  #18 opened 4 years ago by soans1994
  0

• 训练报错

  #5 opened 6 years ago by JeQer
  2

• how to convert these model into coreml model

  #16 opened 4 years ago by Manishsinghrajput98
  0

• 如何在300w数据集上测试？

  #15 opened 4 years ago by zhouqianyu0918
  0

• ValueError: Shape must be rank 0 but is rank 2 for 'loss/output_loss/cond/Switch' (op: 'Switch') with input shapes: [?,166], [?,166].

  #14 opened 4 years ago by jeranio
  0

• How to reproduce this work on another dataset

  #12 opened 5 years ago by bengadisoufiane
  0

• Loading MobileNet has problem perhaps with the version inconsistency

  #11 opened 5 years ago by momo1986
  0

• Getting the pre-trained model for images?

  #10 opened 5 years ago by john-bao-git
  0

• Introduction

  #1 opened 6 years ago by songhengyang
  1

• Image preprocessing

  #9 opened 5 years ago by alexrame
  0

• 请问有没有在数据集上测试过精度之类的？

  #8 opened 5 years ago by keyxuliang
  0

• ....

  #7 opened 5 years ago by cvb2265
  0

• 咨询，该方法速度如何，10ms以内吗？单幅1280*720

  #6 opened 6 years ago by jiagh2010
  2

• Custom dataset and points

  #4 opened 6 years ago by abhigoku10
  9

• 训练的时候报错

  #3 opened 6 years ago by quliulangle
  4

• 请问您使用的TensorFlow和keras的版本？

  #2 opened 6 years ago by quliulangle
  1