YanchaoYang/FDA

Issues

• Simple FDA does not work on CT segmentation.

  #44 opened 5 months ago by mlszy928
  0

• Error in the FDA implementation

  #40 opened 2 years ago by LucasFidon
  2

• Deeplabv2_init question

  #43 opened a year ago by Klopotek0
  0

• Visualization after mean subtraction

  #33 opened 3 years ago by 999Yy999
  3

• Questions about training from scratch

  #8 opened 3 years ago by gustavu92
  2

• DeepLab_init.pth file

  #17 opened 4 years ago by Nadavc220
  1

• Demo.py artifacts in resultant image.

  #42 opened a year ago by Klopotek0
  1

• 3D FDA

  #41 opened a year ago by cong-code
  0

• Question about second SST round

  #12 opened 4 years ago by LeonBre
  5

• synthia_mapped_to_cityscapes(labels)

  #39 opened 3 years ago by laputa-coder
  0

• Target Image pair to Source Image

  #38 opened 3 years ago by BoltenWang-Meta
  0

• Dataset

  #37 opened 3 years ago by zhipengY
  0

• FDA numpy implementation not in line with the paper.

  #36 opened 3 years ago by themurtazanazir
  0

• Slow speed of FDA

  #35 opened 3 years ago by super233
  0

• Model initialization in SSL training rounds

  #34 opened 3 years ago by wangkaihong
  0

• why is the resolution=1344x576

  #24 opened 4 years ago by jh-yi
  1

• Conda Environment specs for running FDA_demo.py + Unable to produce output as the demo

  #32 opened 3 years ago by ukdsvl
  3

• label from synthia has to be convert to CityS format

  #31 opened 3 years ago by 999Yy999
  1

• Why convert to BGR?

  #30 opened 3 years ago by BinhuiXie
  5

• Subtract mean 2 times

  #29 opened 3 years ago by bobvo23
  1

• FDA and random image augmentation methods

  #26 opened 4 years ago by Feobi1999
  1

• ent = ent / 2.9444

  #23 opened 3 years ago by Nadavc220
  1

• initialization weights

  #11 opened 3 years ago by cchenghao
  1

• In fcn8s.py line 162, what does "2.9444" represent? How can I calculate it when I change the number of class? thx

  #9 opened 3 years ago by linlinlin1993
  1

• Questions about FDA.demo.py

  #7 opened 3 years ago by zengkaiwei
  4

• How to get best model of single Beta?

  #6 opened 3 years ago by zengkaiwei
  1

• Problems about Label setting

  #5 opened 3 years ago by liygcheng
  1

• Unable to reproduce the good results of Synthia to Cityscapes

  #22 opened 4 years ago by KaiiZhang
  4

• GPU memory error

  #28 opened 4 years ago by timswim
  1

• label from synthia has to be convert to CityS format

  #27 opened 4 years ago by timswim
  0

• Can we use FDA for faces?

  #25 opened 4 years ago by cvmlddev
  0

• Train from scratch requirements

  #15 opened 4 years ago by Nadavc220
  3

• Question about loss_seg_src in SStrain.py?

  #18 opened 4 years ago by KaiiZhang
  0

• Doubt regarding hardware requirements

  #21 opened 4 years ago by victorvini08
  0

• Question about SYNTHIA dataset classes num

  #20 opened 4 years ago by KaiiZhang
  0

• how to train custom dataset？

  #19 opened 4 years ago by lucafei
  0

• Evaluation checkpoint

  #16 opened 4 years ago by Nadavc220
  0

• Releasing intermediate weights

  #14 opened 4 years ago by akshaykulkarni07
  0

• Domain adaptation

  #13 opened 4 years ago by UralmashFox
  2

• Question about the image transformation process

  #10 opened 4 years ago by liuquande
  8

• Questions about entropy minimization

  #4 opened 4 years ago by jxhuang0508
  2

• Unable to reproduce the results of Sim2Real Adaptation Using FDA (single beta)

  #2 opened 4 years ago by royee182
  9

• Not able to understand this line

  #3 opened 4 years ago by kaushal-py
  1

• looks promising, would love to try out your code!

  #1 opened 4 years ago by zoharbarzelay
  7