clks-wzz/FAS-SGTD

Issues

• How to evaluate on new dataset?

  #9 opened 4 years ago by marziehoghbaie
  14

• evaluating on OULU video frames gave an error

  #28 opened 5 months ago by JavanehBahrami
  0

• Problems when testing.

  #19 opened 4 years ago by eesaeedkarimi
  4

• How to pass the testing samples to the model?

  #27 opened 2 years ago by vuhtr
  0

• problems when tesing

  #18 opened 4 years ago by Xiaofeng-life
  2

• Mask is an array full of 0, leads to Nan for logits

  #22 opened 4 years ago by gabrielvannier
  2

• What should be the names of frame images?

  #26 opened 3 years ago by cuiwenbing
  0

• could you please guide how to download backbone params?

  #25 opened 3 years ago by guoqiangl
  0

• Thank you for your work. Can you provide a single frame pre training model?

  #24 opened 3 years ago by fanchunpeng
  0

• the single_frame in mydata

  #7 opened 4 years ago by zj19921221
  9

• problems when using Oulu_NPU dataset

  #17 opened 4 years ago by Xiaofeng-life
  2

• the shape of input

  #15 opened 4 years ago by wuwuwuyuanhang
  4

• Live video and performance

  #21 opened 4 years ago by OrkhanHI
  0

• convert ckpt to pb

  #20 opened 4 years ago by a5704607
  0

• NaN validation loss while training the phase 1

  #16 opened 4 years ago by ThanhNguyenTung
  2

• Encountered a problem while trying to train on multiple gpus.

  #13 opened 4 years ago by 10LGUO
  2

• Is the dataset DMAD publicly available?

  #14 opened 4 years ago by RizhaoCai
  0

• data and label format description

  #11 opened 4 years ago by ysm022
  1

• Train muti-frame model by using self-data

  #1 opened 4 years ago by CodingMice
  4

• download model

  #6 opened 4 years ago by Medencova
  1

• 数据准备

  #8 opened 4 years ago by huazaikai
  1

• No module named 'tensorflow.contrib'

  #12 opened 4 years ago by 10LGUO
  1

• Dataset download

  #10 opened 4 years ago by ysm022
  1

• Cannot access to baidu, google drive mb?

  #4 opened 4 years ago by cvengneer
  1

• Can you advice on how to modify your example from OULU-NPU to Reply-Attack dataset?

  #5 opened 4 years ago by kai06046
  1

• Conv2_cd block with 3x3 instead of 1x1 convolution

  #3 opened 4 years ago by SNeugber
  2

• Inconsistency with paper w.r.t spatial gradient & 1x1 convolution

  #2 opened 4 years ago by SNeugber
  2