panda-lab/face-landmark

Issues

• 请问数据集我看着是10个关键点，是怎么做到识别60多个的？

  #25 opened 7 years ago by andeyeluguo
  1

• EliteBook 840 G3笔记本，检测和识别耗时700ms

  #37 opened 6 years ago by akfheaven
  5

• 使用opencv加载caffe模型，预测关键点特别不准

  #42 opened 6 years ago by BenJaminKen
  5

• Module face_landmark_localization在移动端比如RK3288上检测出68个关键点需要多长时间？

  #44 opened 4 years ago by hlguo
  0

• 使用ibug，helen，300W数据训练，loss很大，不收敛

  #10 opened 7 years ago by mixml
  10

• 人脸关键点顺序

  #20 opened 7 years ago by reaneyli
  1

• ImportError: No module named python.NormlizedMSE

  #21 opened 7 years ago by freedanfan
  1

• How to make training prediction on one's own data set

  #39 opened 6 years ago by zhouhao-learning
  1

• Would you like to provide a more standard version?

  #28 opened 6 years ago by urbaneman
  1

• 脚本执行流程

  #38 opened 6 years ago by nxf48054
  1

• 如何优化视频中关键点抖动

  #43 opened 5 years ago by zoooo0820
  1

• have a question

  #41 opened 6 years ago by kwangrok92
  3

• Is it right? gary_img = cv2.cvtColor(roi, cv2.COLOR_RGB2GRAY) in face_landmark.py line 65

  #40 opened 6 years ago by DuckJ
  0

• 链接无法下载AFW/ AFLW数据；Can't download the datasets

  #36 opened 6 years ago by angeladzl
  5

• 训练h5文件的对应顺序

  #12 opened 7 years ago by zhanglaplace
  7

• 数据集

  #32 opened 6 years ago by Andrawzyf
  2

• 关于mainloop生成HD5训练数据

  #35 opened 6 years ago by Andrawzyf
  0

• landmark68.txt文件

  #34 opened 6 years ago by Andrawzyf
  0

• 转ＨＤ５格式要花好久

  #33 opened 6 years ago by Andrawzyf
  1

• 关于 train_prototxt

  #30 opened 6 years ago by Andrawzyf
  0

• 关于转换成HD5的问题

  #31 opened 6 years ago by Andrawzyf
  3

• Paper

  #19 opened 6 years ago by td042
  2

• 怎么在Ubuntu编译这个工程呢。。。

  #29 opened 6 years ago by Andrawzyf
  2

• face landmark 训练收敛到face landmark的均值点，怎么解决？

  #27 opened 6 years ago by nan10921p
  0

• 关于如何生成训练数据

  #8 opened 6 years ago by YjDai
  8

• 训练时用的 solver.prototxt 能参考一下吗？

  #26 opened 6 years ago by freedanfan
  0

• 看着数据集只有5个关键点，兄台是怎么做到训练识别60多个的

  #24 opened 7 years ago by andeyeluguo
  1

• where is the absval_layer?

  #23 opened 7 years ago by andeyeluguo
  1

• 眼睛检测不准确

  #22 opened 7 years ago by kenh1991
  1

• 请问训练方法对应的是那一篇论文

  #17 opened 7 years ago by honghuCode
  2

• face landmark 多个点相当于多标签？

  #18 opened 7 years ago by liuqunzhong
  1

• how to add python layer

  #16 opened 7 years ago by ForestWang
  0

• Bug in DataRow

  #15 opened 7 years ago by jasminezz
  0

• scaledLM = dataRow.landmarksScaledMinus05_plus05()请教原理

  #14 opened 7 years ago by jasminezz
  0

• 怎么用dlib才能训练出3M的模型呢？

  #13 opened 7 years ago by jasminezz
  3

• 训练数据

  #6 opened 7 years ago by boundles
  2

• five point

  #4 opened 7 years ago by jmu201521121021
  4

• 是否能提供一份训练数据？

  #11 opened 7 years ago by mixml
  1

• model size

  #9 opened 7 years ago by hellocsi
  2

• How to train ?

  #5 opened 7 years ago by qqning
  2

• train model的caffe solver

  #7 opened 7 years ago by boundles
  3

• 'hsi' is not a member of 'dlib::pixel_traits<int>'

  #3 opened 7 years ago by jiangwqcooler
  7

• How can i test your amazing work?

  #1 opened 7 years ago by jiangwqcooler
  13

• 你好，能给出模型相关方法的原论文吗？

  #2 opened 7 years ago by kuyun-zhangyang
  2