yaojieliu/ECCV20-STDN
Source code for ECCV 2020 paper: On Disentangling Spoof Trace for Generic Face Anti-Spoofing
PythonNOASSERTION
Issues
- 8
what are scores mean?
#18 opened by AtefehBaniasadi - 0
- 2
Query: lm_reverse_list in model/dataset.py has 66 face landmark points specified
#10 opened by GayatriPurandharT - 5
- 2
Calculation of score from score.txt ?
#21 opened by Kapil-23 - 1
Calculation of G(I)
#22 opened by Sonshs - 1
Calculation of pixel wise loss.
#20 opened by MatZlo - 3
Can you release the evaluation code?
#19 opened by Elijah-Yi - 0
SiW video to image
#17 opened by debasmitdas - 1
aboat the label of ESR
#16 opened by Elijah-Yi - 1
BN update problem
#11 opened by chenxinhua - 2
Error load checkpoint from pretrain
#14 opened by ngoanpv - 0
about the model‘s channels
#15 opened by 1relia - 1
can you share the trained weights with me?
#13 opened by JRZNBU - 2
Need suggestion for input
#7 opened by GayatriPurandharT - 0
Preprocess Scripts
#9 opened by carlosbertoncelli - 3
Was use dlib got 68 landmark?
#3 opened by zzzkk2009 - 0
Is it possible to have a demo?
#6 opened by sonnguyenasu - 0
- 0
This is not an issue but I want to ask does the way we crop the face (using different face detectors) affects the prediction of the model?
#2 opened by xxxpsyduck