clovaai/ext_portrait_segmentation

Issues

• about test！！！

  #27 opened 4 years ago by pantingchn
  4

• custom dataset training

  #35 opened 2 years ago by guoqiushi
  1

• Pretrained model output bad segmentation result

  #31 opened 4 years ago by edvardHua
  4

• trainning new data and test error, help please

  #32 opened 4 years ago by zhangyunming
  3

• Accuracy issue, implementation difference compared to paper.

  #17 opened 5 years ago by avinash31d
  4

• groups config setting

  #34 opened 4 years ago by Zachary-ZCH
  0

• Model does not perform well on real world portrait images/videos

  #33 opened 4 years ago by anilsathyan7
  1

• Error in Visualize_video.py when using result/SINet/SINet.pth

  #25 opened 4 years ago by Novaal
  5

• pickle file in the read me file for custom dataset

  #4 opened 5 years ago by freaad
  8

• Converting .pth model to onnx

  #28 opened 4 years ago by richard-salaschavez
  7

• why ignore_idx==255? in lovasz loss?

  #30 opened 4 years ago by Minotaur-CN
  2

• Configuration for training baidu_V2

  #1 opened 4 years ago
  5

• ValueError: low >= high

  #8 opened 5 years ago by dengfenglai321
  6

• visualizing results

  #18 opened 5 years ago by ucb-pb
  4

• about inference time

  #29 opened 4 years ago by HeaderZzj
  0

• Why is the label of eg1800 data set inconsistent with the number of images?

  #26 opened 5 years ago by TCBocean
  3

• Pretrained model

  #2 opened 5 years ago
  12

• The speed about ExtremeC3Net.

  #19 opened 5 years ago by Med-Process
  1

• How to use EG1800

  #24 opened 5 years ago by ohheysherry66
  0

• main.py

  #22 opened 5 years ago by ZshahRA
  1

• syn_bg.jpg

  #6 opened 5 years ago by jxust01
  3

• Testing new Image

  #23 opened 5 years ago by ZshahRA
  2

• Rough edges in mask?

  #20 opened 5 years ago by InternetMaster1
  2

• Training new dataset

  #21 opened 5 years ago by ZshahRA
  0

• About the implement of information bolcking

  #12 opened 5 years ago by Almostbealone
  1

• where is 'syn_bg.jpg'? it seem not right

  #15 opened 5 years ago by huangzicheng
  1

• the size of input_image

  #16 opened 5 years ago by luqinghit
  1

• how to control the channel output to match lovasz vs cross entropy loss?

  #14 opened 5 years ago by lessw2020
  1

• training tip

  #10 opened 5 years ago by freaad
  4

• TypeError: '>' not supported between instances of 'NoneType' and 'int'

  #13 opened 5 years ago by dengfenglai321
  1

• can't download dataset

  #11 opened 5 years ago by comihuang1030
  1

• ValueError: invalid literal for int() with base 10: '501.png\n'

  #7 opened 5 years ago by dengfenglai321
  1