Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Unconditional Watermarked Generation on AFHQ 64x64 v2 (train)
Loading...
2.1
FID (50k Samples)
MDM-proj
1.9124
3.1787
4.445
5.7113
Oct 2, 2023
FID (50k Samples)
FID* (50k Samples)
Updated 1mo ago
Evaluation Results
Method
Method
Links
FID (50k Samples)
FID* (50k Samples)
MDM-proj
Precision=56.9%
2023.10
2.1
-
MDM-proj
Precision=75.0%
2023.10
2.12
-
MDM-dual
Precision=56.9%
2023.10
2.23
2.21
MDM-proj
Precision=92.7%
2023.10
2.3
-
MDM-dual
Precision=75.0%
2023.10
2.86
2.32
Prior watermarked diffusion model
2023.10
4.32
-
MDM-dual
Precision=92.7%
2023.10
6.79
3.05
Feedback
Search any
task
Search any
task