少女祈祷中...
Home
Archives
About
Friend
Diffusion-Backdoor-Embed
2024-04-18
实验
Hey, password is required here.
Newer
TrojDiff
Older
VDC
Contents
1.
idea
1.1.
背景
1.2.
想法
1.3.
威胁模型
1.4.
实验
2.
代码
2.1.
loss1
2.2.
loss2
2.3.
loss3
2.4.
loss4
2.5.
loss5
3.
实验 – pred x_0
3.1.
benign diffusion model
3.1.1.
res_benign_cifar10_step1k
3.1.2.
res_benign_cifar10_step10k
3.1.3.
res_benign_cifar10_step15k -> lab
3.2.
bad diffusion model dataset error
3.2.1.
res_badnet_grid_cifar10_step1k_ratio2_loss1
3.2.2.
res_badnet_grid_cifar10_step1k_ratio5_loss_1
3.2.3.
res_badnet_grid_cifar10_step10k_ratio2_loss1
3.2.4.
res_badnet_grid_cifar10_step10k_ratio2_loss2
3.2.5.
res_badnet_grid_cifar10_step10k_ratio2_loss3
3.2.6.
res_badnet_grid_cifar10_step10k_ratio2_loss4
3.2.6.1.
factor1 [2, 2, 6]
3.2.6.2.
factor2 [4, 3, 6]
3.2.6.3.
factor3 [3, 3, 1]
3.2.7.
res_badnet_grid_cifar10_step10k_ratio1_loss4
3.2.7.1.
factor1 [3, 3, 1]
3.2.7.2.
factor2 [2, 2, 0.7]
3.2.7.3.
factor3 [2, 2, 0.1]
3.2.8.
res_badnet_grid_cifar10_step10k_ratio1_loss5
3.2.8.1.
factor1 [2, 2, 0.1]
3.2.8.2.
factor2 [2, 2, 0.7]
3.3.
bad diffusion model devide1:5
3.3.1.
res_badnet_grid_cifar10_step15k_ratio1_loss5
3.3.1.1.
factor1 [2, 2, 0.3] -> pc
3.3.1.2.
factor1 [2, 2, 0.5] -> lv-pc
3.4.
bad diffusion model divide1:9
3.4.1.
res_badnet_grid_cifar10_part10_step15k_ratio1_loss5
3.4.1.1.
factor1 [2, 2, 0.1] -> pc
3.4.1.2.
factor2 [2, 2, 1] -> lv-pc
3.5.
bad diffusion model DatasetMix_1
3.5.1.
res_badnet_grid_cifar10_step15k_ratio1_loss5
3.5.1.1.
factor1 [2, 2, 0.5] -> lv
3.5.1.2.
factor2 [2, 2, 1] -> lab
3.6.
bad diffusion model DatasetMix_2
3.6.1.
res_badnet_grid_cifar10_step15k_ratio1_loss5
3.6.1.1.
factor1 [1, 1, 1] -> lv
3.6.1.2.
factor2 [1, 1, 2] -> pc
3.7.
bad diffusion model DatasetMix_3
3.7.1.
try1 -> pc
3.7.2.
try2 -> lv
3.7.3.
try3 -> lab
3.7.4.
try4 -> lv
3.7.5.
try5 -> pc
3.7.6.
try6 -> lab
3.7.7.
try7 -> lv
3.8.
bad diffusion model DatasetMix_4
3.8.1.
mf1 -> pc
3.8.2.
mf2 -> pc
3.8.3.
mf3 -> lab
3.8.4.
mf4 -> lab
3.8.5.
mf5 -> pc
3.8.6.
mf6 -> pc
3.8.7.
mf7 -> lv
4.
实验 – pred noice
4.1.
res_exp1 -> lv
4.2.
res_exp2 -> pc
4.3.
res_exp3 -> lab
4.4.
res_exp4 -> lv
4.5.
res_exp5 -> pc
4.6.
res_exp_6 -> lv
4.7.
res_exp7 -> lv
5.
实验
5.1.
res_exp1 -> lv
5.2.
res_exp2 -> lab
5.3.
res_exp3 -> lv
5.4.
res_exp4 -> lab
5.5.
res_exp5 -> lv
5.6.
res-exp6 -> pc
5.7.
res_exp7 -> lab
5.8.
res_exp8 -> lv
5.9.
res_exp9 -> pc
5.10.
res_exp10 -> pc
5.11.
res_exp11 -> pc
5.12.
res_exp12 -> lv
5.13.
res_exp13 -> pc
5.14.
res_exp14 -> pc
5.15.
res_exp15 -> pc
5.16.
res_exp16 -> lab
5.17.
res_exp17 -> pc
5.18.
res_exp18 -> lab
5.19.
res_exp19 -> lab
6.
blended
6.1.
imagenette
6.1.1.
exp2
6.2.
GTSRB
6.2.1.
exp1
6.2.2.
exp2
7.
badnet
7.1.
cifar10
7.1.1.
exp_1
8.
benign
8.1.
imagenette
9.
reference
10.
ASR
chengyiqiu
Posts
65
Categories
13
Tags
17
Home
Archives
About
Friend
