Could you share a train log? #6
Comments
|
Sorry that I cannot get my hard disk in university with previous results now because of the COVID19. But I can share you a similar log (Another method not CDCN++, but similar): Oulu-NPU, P1: epoch:280, Val: val_threshold= 0.0014, val_ACC= 0.9689, val_ACER= 0.0320 epoch:281, Train: Absolute_Depth_loss= 0.0134, Contrastive_Depth_loss= 0.0044 epoch:300, Val: val_threshold= 0.0014, val_ACC= 0.9811, val_ACER= 0.0202 epoch:301, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0043 epoch:320, Val: val_threshold= 0.0015, val_ACC= 0.9689, val_ACER= 0.0320 epoch:321, Train: Absolute_Depth_loss= 0.0104, Contrastive_Depth_loss= 0.0040 epoch:340, Val: val_threshold= 0.0015, val_ACC= 0.9755, val_ACER= 0.0257 epoch:341, Train: Absolute_Depth_loss= 0.0101, Contrastive_Depth_loss= 0.0042 epoch:360, Val: val_threshold= 0.0016, val_ACC= 0.9811, val_ACER= 0.0202 epoch:361, Train: Absolute_Depth_loss= 0.0114, Contrastive_Depth_loss= 0.0042 epoch:380, Val: val_threshold= 0.0019, val_ACC= 0.9811, val_ACER= 0.0202 epoch:381, Train: Absolute_Depth_loss= 0.0094, Contrastive_Depth_loss= 0.0039 epoch:400, Val: val_threshold= 0.0020, val_ACC= 0.9711, val_ACER= 0.0306 epoch:401, Train: Absolute_Depth_loss= 0.0098, Contrastive_Depth_loss= 0.0040 epoch:420, Val: val_threshold= 0.0013, val_ACC= 0.9811, val_ACER= 0.0202 epoch:421, Train: Absolute_Depth_loss= 0.0094, Contrastive_Depth_loss= 0.0038 epoch:440, Val: val_threshold= 0.0013, val_ACC= 0.9811, val_ACER= 0.0202 epoch:441, Train: Absolute_Depth_loss= 0.0094, Contrastive_Depth_loss= 0.0039 epoch:460, Val: val_threshold= 0.0017, val_ACC= 0.9822, val_ACER= 0.0195 epoch:461, Train: Absolute_Depth_loss= 0.0093, Contrastive_Depth_loss= 0.0037 epoch:480, Val: val_threshold= 0.0016, val_ACC= 0.9867, val_ACER= 0.0146 epoch:481, Train: Absolute_Depth_loss= 0.0078, Contrastive_Depth_loss= 0.0037 epoch:500, Val: val_threshold= 0.0020, val_ACC= 0.9844, val_ACER= 0.0160 epoch:501, Train: Absolute_Depth_loss= 0.0068, Contrastive_Depth_loss= 0.0034 epoch:520, Val: val_threshold= 0.0019, val_ACC= 0.9878, val_ACER= 0.0139 epoch:521, Train: Absolute_Depth_loss= 0.0057, Contrastive_Depth_loss= 0.0033 epoch:540, Val: val_threshold= 0.0015, val_ACC= 0.9878, val_ACER= 0.0139 epoch:541, Train: Absolute_Depth_loss= 0.0051, Contrastive_Depth_loss= 0.0033 epoch:560, Val: val_threshold= 0.0019, val_ACC= 0.9889, val_ACER= 0.0132 epoch:561, Train: Absolute_Depth_loss= 0.0055, Contrastive_Depth_loss= 0.0033 epoch:580, Val: val_threshold= 0.0021, val_ACC= 0.9878, val_ACER= 0.0118 epoch:581, Train: Absolute_Depth_loss= 0.0052, Contrastive_Depth_loss= 0.0032 epoch:600, Val: val_threshold= 0.0020, val_ACC= 0.9878, val_ACER= 0.0118 epoch:601, Train: Absolute_Depth_loss= 0.0053, Contrastive_Depth_loss= 0.0031 epoch:620, Val: val_threshold= 0.0015, val_ACC= 0.9867, val_ACER= 0.0146 epoch:621, Train: Absolute_Depth_loss= 0.0051, Contrastive_Depth_loss= 0.0031 epoch:640, Val: val_threshold= 0.0016, val_ACC= 0.9855, val_ACER= 0.0153 epoch:641, Train: Absolute_Depth_loss= 0.0047, Contrastive_Depth_loss= 0.0031 epoch:660, Val: val_threshold= 0.0021, val_ACC= 0.9867, val_ACER= 0.0146 epoch:661, Train: Absolute_Depth_loss= 0.0045, Contrastive_Depth_loss= 0.0031 epoch:680, Val: val_threshold= 0.0016, val_ACC= 0.9867, val_ACER= 0.0146 epoch:681, Train: Absolute_Depth_loss= 0.0047, Contrastive_Depth_loss= 0.0031 epoch:700, Val: val_threshold= 0.0019, val_ACC= 0.9822, val_ACER= 0.0174 epoch:701, Train: Absolute_Depth_loss= 0.0047, Contrastive_Depth_loss= 0.0029 epoch:720, Val: val_threshold= 0.0015, val_ACC= 0.9867, val_ACER= 0.0146 epoch:721, Train: Absolute_Depth_loss= 0.0045, Contrastive_Depth_loss= 0.0030 epoch:740, Val: val_threshold= 0.0015, val_ACC= 0.9867, val_ACER= 0.0125 epoch:741, Train: Absolute_Depth_loss= 0.0048, Contrastive_Depth_loss= 0.0031 epoch:760, Val: val_threshold= 0.0014, val_ACC= 0.9867, val_ACER= 0.0146 epoch:761, Train: Absolute_Depth_loss= 0.0046, Contrastive_Depth_loss= 0.0030 epoch:780, Val: val_threshold= 0.0016, val_ACC= 0.9867, val_ACER= 0.0146 epoch:781, Train: Absolute_Depth_loss= 0.0042, Contrastive_Depth_loss= 0.0030 epoch:800, Val: val_threshold= 0.0022, val_ACC= 0.9833, val_ACER= 0.0188 epoch:801, Train: Absolute_Depth_loss= 0.0047, Contrastive_Depth_loss= 0.0031 epoch:820, Val: val_threshold= 0.0012, val_ACC= 0.9855, val_ACER= 0.0153 epoch:821, Train: Absolute_Depth_loss= 0.0040, Contrastive_Depth_loss= 0.0028 epoch:840, Val: val_threshold= 0.0020, val_ACC= 0.9878, val_ACER= 0.0139 epoch:841, Train: Absolute_Depth_loss= 0.0042, Contrastive_Depth_loss= 0.0029 epoch:860, Val: val_threshold= 0.0016, val_ACC= 0.9867, val_ACER= 0.0146 epoch:861, Train: Absolute_Depth_loss= 0.0053, Contrastive_Depth_loss= 0.0030 epoch:880, Val: val_threshold= 0.0019, val_ACC= 0.9855, val_ACER= 0.0153 epoch:881, Train: Absolute_Depth_loss= 0.0048, Contrastive_Depth_loss= 0.0030 epoch:900, Val: val_threshold= 0.0018, val_ACC= 0.9867, val_ACER= 0.0146 epoch:901, Train: Absolute_Depth_loss= 0.0042, Contrastive_Depth_loss= 0.0028 epoch:920, Val: val_threshold= 0.0017, val_ACC= 0.9878, val_ACER= 0.0139 epoch:921, Train: Absolute_Depth_loss= 0.0040, Contrastive_Depth_loss= 0.0029 epoch:940, Val: val_threshold= 0.0018, val_ACC= 0.9822, val_ACER= 0.0195 epoch:941, Train: Absolute_Depth_loss= 0.0043, Contrastive_Depth_loss= 0.0029 epoch:960, Val: val_threshold= 0.0018, val_ACC= 0.9855, val_ACER= 0.0153 epoch:961, Train: Absolute_Depth_loss= 0.0043, Contrastive_Depth_loss= 0.0028 epoch:980, Val: val_threshold= 0.0012, val_ACC= 0.9867, val_ACER= 0.0146 epoch:981, Train: Absolute_Depth_loss= 0.0040, Contrastive_Depth_loss= 0.0028 epoch:1000, Val: val_threshold= 0.0015, val_ACC= 0.9878, val_ACER= 0.0139 epoch:1001, Train: Absolute_Depth_loss= 0.0033, Contrastive_Depth_loss= 0.0027 epoch:1020, Val: val_threshold= 0.0018, val_ACC= 0.9867, val_ACER= 0.0146 |
|
Thank your reply! I try again. |
|
@xiao-keeplearning Hi, how much time you waited for to get oulu-npu dataset after signing the license agreement? Thanks in advance |
|
Hi，Can you share me the dataset，I signed the agreement but did not reply。 |
About a week. |
You can leave your email, I will share it with you. |
|
@xiao-keeplearning can I ask you to share with me? I am having same trouble. This is my mail cvengneer@gmail.com. Thank you a lot! |
@xiao-keeplearning Hi, I applied about a month back but no reply from them. I need the dataset for my recent research, can you share it with me please ? Will be much helpful. |
|
@xiao-keeplearning 您好，想问下您现在训练时ACER稳定嘛，我在oulu上训练时ACER浮动也比较大，没有作者的稳定，结果也没有论文里的好。如果您的效果变稳定了，可以告诉我下您进行了哪些修改嘛，十分困扰，谢谢。期待您的回复！我的邮箱：1421159991@qq.com |
|
@lz28lz28 It can be seen from the log given by the Owner that ACER on the test set fluctuates during training.
I have reproduced the paper result. In my opinion, ACER is confusing, you can first focus on ACC. When Acc reaches 0.9967(Random size crop is important, you can easily reach it), it means 2 samples were misjudged. As we know, for Proto1. ,
Then, you can adjust the crop scale for face in the test phase to search the best scale which makes misjudge examples all belong to fake faces. At last, you will get the paper results. |
感谢您的指导，i will try again. |
您好 我也是用学生邮箱提交了申请但是没有收到回复，您能否也分享一份给我呢？感激不尽！这是我的邮箱g20198868@xs.ustb.edu.cn |
|
@xiao-keeplearning Can you share me the dataset，I signed the agreement but it did not reply。 My email is lalaland19704@gmail.com |
Can you share me the dataset，Thank you very much. My email is shengxin@nwafu.edu.cn |
Hi，my email is 2294556197@qq.com, could you share with me? |
|
@Prasandhmcw @xiao-keeplearning @EchoIR @lz28lz28 @silvercherry FileNotFoundError: [WinError 3] The system cannot find the path specified: 'D:/face_anti_spoofing/dataset/OULUNPU/Data/demo/Train/Depth_map_1/1_1_01_3' Even I have images in this folder,... How about the def single_iamge_x(): I have confusion in it, Because i have already splitted up the oulu dataset videos 1_1_01_1 into 1_1_01_1_1, 1_1_01_1_2,1_1_01_1_3 like this...., def get_single_image_x(self, image_path, map_path, videoname): here the error occurs, FileNotFoundError: [WinError 3] The system cannot find the path specified: 'D:/face_anti_spoofing/dataset/OULUNPU/Data/demo/Train/Depth_map_1/1_1_01_3',.. any to please, explain to me about it,. Thanks in advance |
|
Hi @xiao-keeplearning and all, I have signed OULUNPU dataset but dont received response yet. |
@xiao-keeplearning Hi, I also did not have a response for applying for the agreement. Can you share me the dataset? My email is marvelyou.fan@gmail.com. Thank you! . |
Hi，my email is 1334637558@qq.com, could you share with me? |
|
@xiao-keeplearning 大佬可以分享一下数据集吗，申请了一周没有得到回复，我的邮箱是460579132@qq.com 谢谢 |
|
@xiao-keeplearning Can I ask you to share the dataset with me? I haven't receive any updates to my dataset request for about two weeks. This is my mail: dang.nh160988@sis.hust.edu.vn . Thank you a lot! |
Hi! I also signed the agreement but didn't get any response yet. |
|
@xiao-keeplearning Could you kindly share the preprocessing steps you took or maybe share the scripts that you used here? |
|
HI,
You can use PRNet to generate the depth map, and MTCNN for boundary box
detection @seyyed Hossein Hasanpour
…On Mon, 24 May 2021 at 13:23, Seyyed Hossein Hasanpour < ***@***.***> wrote:
@xiao-keeplearning <https://github.com/xiao-keeplearning> Could you
kindly share the preprocessing steps you took or maybe share the scripts
that you used here?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B26K4RDFPVLMESG724LTPHPEPANCNFSM4LVPHJTA>
.
|
|
@punitha-valli Thanks a lot. really appreciate it. what about the frame extractions? at what interval should I be extracting frames from the videos? |
|
It's depends upon your priority
You can use all the frames of the video or else you can use only the
selective frames
Hope good for your research
…On Mon, May 24, 2021, 2:13 PM Seyyed Hossein Hasanpour < ***@***.***> wrote:
@punitha-valli <https://github.com/punitha-valli> Thanks a lot. really
appreciate it. what about the frame extractions? at what interval should I
be extracting frames from the videos?
And for PRNet which repo was used exactly?
What about the MTCNNs?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B252CHV4WOWJIXK4JXLTPHU7NANCNFSM4LVPHJTA>
.
|
|
Thanks, my first priority is to replicate the exact results this paper achieves, and then build on top of that. without knowing the exact configuration used, I can not compare my work with them or be certain we are comparing apples to apples and not oranges. |
|
Hi @punitha-valli , how long you trained 1 epoch? I trained 1 epoch on GPU P100 spending 10 hours, is it normal? |
|
HI,
I use NVIDIA 1080 , it took nearly 15 min for one epoch,
please check your code about the GPU, i mean the parameters passed into the
GPU
MY best wishes for your work
…On Thu, 27 May 2021 at 14:48, Pham Van Ngoan ***@***.***> wrote:
Hi @punitha-valli <https://github.com/punitha-valli> , how long you
trained 1 epoch? I trained 1 epoch on GPU P100 spending 10 hours, is it
normal?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B26FXR3R3ZK7HWPFK7TTPXTNTANCNFSM4LVPHJTA>
.
|
|
|
@Coderx7 @ngoanpv @marvelyou @ngoanpv @22wei22 @Prasandhmcw @ZitongYu @lz28lz28 @GuoQuanhao @lz28lz28 @laoxing797373 @luan1412167 @dangnh0611 @xiao-keeplearning I have used the same code and OULU NPU dataset, my loss values during training are the same with the author's as he mentioned in the previous But my ACER value is very very high I have attached my file for reference Can anyone please help me? tit will be a great help Oulu-NPU, P1: epoch:1, Val: val_threshold= 0.0003, val_ACC= 0.4341, val_ACER= 0.5731 epoch:21, Val: val_threshold= 0.0007, val_ACC= 0.5659, val_ACER= 0.4511 epoch:41, Val: val_threshold= 0.0008, val_ACC= 0.4806, val_ACER= 0.5265 epoch:61, Val: val_threshold= 0.0008, val_ACC= 0.5504, val_ACER= 0.4556 epoch:81, Val: val_threshold= 0.0009, val_ACC= 0.4884, val_ACER= 0.5182 epoch:101, Val: val_threshold= 0.0008, val_ACC= 0.5116, val_ACER= 0.5040 epoch:121, Val: val_threshold= 0.0008, val_ACC= 0.5039, val_ACER= 0.5036 epoch:141, Val: val_threshold= 0.0006, val_ACC= 0.4574, val_ACER= 0.5542 epoch:161, Val: val_threshold= 0.0009, val_ACC= 0.4496, val_ACER= 0.5664 epoch:181, Val: val_threshold= 0.0008, val_ACC= 0.3953, val_ACER= 0.5943 epoch:201, Val: val_threshold= 0.0007, val_ACC= 0.5504, val_ACER= 0.4477 epoch:221, Val: val_threshold= 0.0009, val_ACC= 0.5891, val_ACER= 0.4292 epoch:241, Val: val_threshold= 0.0007, val_ACC= 0.5969, val_ACER= 0.4047 epoch:261, Val: val_threshold= 0.0005, val_ACC= 0.5426, val_ACER= 0.4660 epoch:281, Val: val_threshold= 0.0007, val_ACC= 0.6047, val_ACER= 0.4019 epoch:301, Val: val_threshold= 0.0008, val_ACC= 0.5659, val_ACER= 0.4401 epoch:321, Val: val_threshold= 0.0009, val_ACC= 0.6589, val_ACER= 0.3960 epoch:341, Val: val_threshold= 0.0007, val_ACC= 0.5891, val_ACER= 0.4202 epoch:361, Val: val_threshold= 0.0010, val_ACC= 0.5116, val_ACER= 0.5026 epoch:381, Val: val_threshold= 0.0010, val_ACC= 0.6047, val_ACER= 0.4141 epoch:401, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4879 epoch:421, Val: val_threshold= 0.0006, val_ACC= 0.5581, val_ACER= 0.4469 epoch:441, Val: val_threshold= 0.0006, val_ACC= 0.4884, val_ACER= 0.5074 epoch:461, Val: val_threshold= 0.0005, val_ACC= 0.5659, val_ACER= 0.4511 epoch:481, Val: val_threshold= 0.0004, val_ACC= 0.4729, val_ACER= 0.5331 epoch:501, Val: val_threshold= 0.0008, val_ACC= 0.6202, val_ACER= 0.3919 epoch:521, Val: val_threshold= 0.0006, val_ACC= 0.5194, val_ACER= 0.4804 epoch:541, Val: val_threshold= 0.0005, val_ACC= 0.4884, val_ACER= 0.5132 epoch:561, Val: val_threshold= 0.0005, val_ACC= 0.5271, val_ACER= 0.4886 epoch:581, Val: val_threshold= 0.0005, val_ACC= 0.5116, val_ACER= 0.5014 epoch:601, Val: val_threshold= 0.0007, val_ACC= 0.5039, val_ACER= 0.5086 epoch:621, Val: val_threshold= 0.0006, val_ACC= 0.5271, val_ACER= 0.4839 epoch:641, Val: val_threshold= 0.0005, val_ACC= 0.5039, val_ACER= 0.4975 epoch:661, Val: val_threshold= 0.0008, val_ACC= 0.6279, val_ACER= 0.3868 epoch:681, Val: val_threshold= 0.0005, val_ACC= 0.6434, val_ACER= 0.3637 epoch:701, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4871 epoch:721, Val: val_threshold= 0.0006, val_ACC= 0.5194, val_ACER= 0.4876 epoch:741, Val: val_threshold= 0.0007, val_ACC= 0.4884, val_ACER= 0.5259 epoch:761, Val: val_threshold= 0.0007, val_ACC= 0.5194, val_ACER= 0.4681 epoch:781, Val: val_threshold= 0.0006, val_ACC= 0.6124, val_ACER= 0.3849 epoch:801, Val: val_threshold= 0.0003, val_ACC= 0.5969, val_ACER= 0.4325 epoch:821, Val: val_threshold= 0.0004, val_ACC= 0.5814, val_ACER= 0.4420 epoch:841, Val: val_threshold= 0.0004, val_ACC= 0.5426, val_ACER= 0.4662 epoch:861, Val: val_threshold= 0.0005, val_ACC= 0.4496, val_ACER= 0.5631 epoch:881, Val: val_threshold= 0.0006, val_ACC= 0.4961, val_ACER= 0.5168 epoch:901, Val: val_threshold= 0.0004, val_ACC= 0.4729, val_ACER= 0.5457 epoch:921, Val: val_threshold= 0.0007, val_ACC= 0.5581, val_ACER= 0.4561 epoch:941, Val: val_threshold= 0.0003, val_ACC= 0.4341, val_ACER= 0.5713 epoch:961, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4883 epoch:981, Val: val_threshold= 0.0006, val_ACC= 0.5116, val_ACER= 0.5058 epoch:1001, Val: val_threshold= 0.0007, val_ACC= 0.5349, val_ACER= 0.4638 epoch:1021, Val: val_threshold= 0.0004, val_ACC= 0.5426, val_ACER= 0.4660 epoch:1041, Val: val_threshold= 0.0004, val_ACC= 0.4806, val_ACER= 0.5293 epoch:1061, Val: val_threshold= 0.0004, val_ACC= 0.5504, val_ACER= 0.4516 epoch:1081, Val: val_threshold= 0.0005, val_ACC= 0.4651, val_ACER= 0.5411 epoch:1101, Val: val_threshold= 0.0004, val_ACC= 0.5504, val_ACER= 0.4763 epoch:1121, Val: val_threshold= 0.0005, val_ACC= 0.5271, val_ACER= 0.4827 epoch:1141, Val: val_threshold= 0.0003, val_ACC= 0.5659, val_ACER= 0.4435 epoch:1161, Val: val_threshold= 0.0003, val_ACC= 0.5969, val_ACER= 0.4171 epoch:1181, Val: val_threshold= 0.0003, val_ACC= 0.5659, val_ACER= 0.4511 epoch:1201, Val: val_threshold= 0.0004, val_ACC= 0.5039, val_ACER= 0.4924 epoch:1221, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4797 epoch:1241, Val: val_threshold= 0.0004, val_ACC= 0.5194, val_ACER= 0.4928 epoch:1261, Val: val_threshold= 0.0005, val_ACC= 0.5659, val_ACER= 0.4435 epoch:1281, Val: val_threshold= 0.0004, val_ACC= 0.5271, val_ACER= 0.4756 |
|
老哥可以给我发下OULU-NPU数据集吗，申请了还没通过0.O
…------------------ 原始邮件 ------------------
发件人: "ZitongYu/CDCN" ***@***.***>;
发送时间: 2021年5月31日(星期一) 下午3:20
***@***.***>;
***@***.******@***.***>;
主题: Re: [ZitongYu/CDCN] Could you share a train log? (#6)
@Coderx7 @ngoanpv @marvelyou @ngoanpv @22wei22 @Prasandhmcw @ZitongYu @lz28lz28 @GuoQuanhao @lz28lz28 @laoxing797373 @luan1412167 @dangnh0611
I have used the same code and OULU NPU dataset,
my loss values during training are the same with the author's as he mentioned in the previous
But my ACER value is very very high
I have attached my file for reference
Can anyone please help me?
tit will be a great help
Oulu-NPU, P1:
train from scratch!
epoch:1, Train: Absolute_Depth_loss= 0.0552, Contrastive_Depth_loss= 0.0072
epoch:1, Val: val_threshold= 0.0003, val_ACC= 0.4341, val_ACER= 0.5731
epoch:1, Test: ACC= 0.1279, APCER= 0.9740, BPCER= 0.0000, ACER= 0.4870
epoch:2, Train: Absolute_Depth_loss= 0.0532, Contrastive_Depth_loss= 0.0059
epoch:3, Train: Absolute_Depth_loss= 0.0512, Contrastive_Depth_loss= 0.0059
epoch:4, Train: Absolute_Depth_loss= 0.0491, Contrastive_Depth_loss= 0.0055
epoch:5, Train: Absolute_Depth_loss= 0.0483, Contrastive_Depth_loss= 0.0052
epoch:6, Train: Absolute_Depth_loss= 0.0479, Contrastive_Depth_loss= 0.0049
epoch:7, Train: Absolute_Depth_loss= 0.0479, Contrastive_Depth_loss= 0.0049
epoch:8, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0048
epoch:9, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0048
epoch:10, Train: Absolute_Depth_loss= 0.0479, Contrastive_Depth_loss= 0.0048
epoch:11, Train: Absolute_Depth_loss= 0.0474, Contrastive_Depth_loss= 0.0048
epoch:12, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0048
epoch:13, Train: Absolute_Depth_loss= 0.0480, Contrastive_Depth_loss= 0.0048
epoch:14, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0047
epoch:15, Train: Absolute_Depth_loss= 0.0474, Contrastive_Depth_loss= 0.0047
epoch:16, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0047
epoch:17, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0047
epoch:18, Train: Absolute_Depth_loss= 0.0472, Contrastive_Depth_loss= 0.0047
epoch:19, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0047
epoch:20, Train: Absolute_Depth_loss= 0.0472, Contrastive_Depth_loss= 0.0047
epoch:21, Train: Absolute_Depth_loss= 0.0472, Contrastive_Depth_loss= 0.0047
epoch:21, Val: val_threshold= 0.0007, val_ACC= 0.5659, val_ACER= 0.4511
epoch:21, Test: ACC= 0.5581, APCER= 0.3167, BPCER= 0.7308, ACER= 0.5237
epoch:22, Train: Absolute_Depth_loss= 0.0472, Contrastive_Depth_loss= 0.0047
epoch:23, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0047
epoch:24, Train: Absolute_Depth_loss= 0.0476, Contrastive_Depth_loss= 0.0047
epoch:25, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0046
epoch:26, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0047
epoch:27, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0047
epoch:28, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:29, Train: Absolute_Depth_loss= 0.0477, Contrastive_Depth_loss= 0.0047
epoch:30, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0046
epoch:31, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0046
epoch:32, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:33, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:34, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:35, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0046
epoch:36, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:37, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:38, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:39, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:40, Train: Absolute_Depth_loss= 0.0472, Contrastive_Depth_loss= 0.0046
epoch:41, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0046
epoch:41, Val: val_threshold= 0.0008, val_ACC= 0.4806, val_ACER= 0.5265
epoch:41, Test: ACC= 0.5000, APCER= 0.4861, BPCER= 0.5714, ACER= 0.5288
epoch:42, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0046
epoch:43, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:44, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:45, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0046
epoch:46, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0046
epoch:47, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:48, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:49, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:50, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0046
epoch:51, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0046
epoch:52, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:53, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:54, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:55, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:56, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:57, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0046
epoch:58, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:59, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:60, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:61, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0046
epoch:61, Val: val_threshold= 0.0008, val_ACC= 0.5504, val_ACER= 0.4556
epoch:61, Test: ACC= 0.6163, APCER= 0.3571, BPCER= 0.5000, ACER= 0.4286
epoch:62, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:63, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:64, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0046
epoch:65, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:66, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:67, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0046
epoch:68, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:69, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:70, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:71, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:72, Train: Absolute_Depth_loss= 0.0471, Contrastive_Depth_loss= 0.0046
epoch:73, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0046
epoch:74, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:75, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:76, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0046
epoch:77, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0046
epoch:78, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0046
epoch:79, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:80, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:81, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:81, Val: val_threshold= 0.0009, val_ACC= 0.4884, val_ACER= 0.5182
epoch:81, Test: ACC= 0.4767, APCER= 0.5278, BPCER= 0.5000, ACER= 0.5139
epoch:82, Train: Absolute_Depth_loss= 0.0475, Contrastive_Depth_loss= 0.0046
epoch:83, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0046
epoch:84, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:85, Train: Absolute_Depth_loss= 0.0473, Contrastive_Depth_loss= 0.0046
epoch:86, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0046
epoch:87, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:88, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:89, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:90, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0046
epoch:91, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:92, Train: Absolute_Depth_loss= 0.0474, Contrastive_Depth_loss= 0.0046
epoch:93, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:94, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:95, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:96, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:97, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:98, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0046
epoch:99, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:100, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:101, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0045
epoch:101, Val: val_threshold= 0.0008, val_ACC= 0.5116, val_ACER= 0.5040
epoch:101, Test: ACC= 0.6628, APCER= 0.2838, BPCER= 0.6667, ACER= 0.4752
epoch:102, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:103, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:104, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:105, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:106, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:107, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:108, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:109, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:110, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:111, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0046
epoch:112, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:113, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0045
epoch:114, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:115, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:116, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:117, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:118, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:119, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:120, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:121, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:121, Val: val_threshold= 0.0008, val_ACC= 0.5039, val_ACER= 0.5036
epoch:121, Test: ACC= 0.5116, APCER= 0.4531, BPCER= 0.5909, ACER= 0.5220
epoch:122, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0045
epoch:123, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:124, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:125, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:126, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0045
epoch:127, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:128, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:129, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0045
epoch:130, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0045
epoch:131, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:132, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:133, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:134, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:135, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:136, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:137, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:138, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:139, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0045
epoch:140, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:141, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0045
epoch:141, Val: val_threshold= 0.0006, val_ACC= 0.4574, val_ACER= 0.5542
epoch:141, Test: ACC= 0.5233, APCER= 0.4853, BPCER= 0.4444, ACER= 0.4649
epoch:142, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:143, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:144, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0045
epoch:145, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:146, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:147, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:148, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:149, Train: Absolute_Depth_loss= 0.0456, Contrastive_Depth_loss= 0.0045
epoch:150, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:151, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:152, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:153, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:154, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:155, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:156, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:157, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:158, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0045
epoch:159, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:160, Train: Absolute_Depth_loss= 0.0468, Contrastive_Depth_loss= 0.0045
epoch:161, Train: Absolute_Depth_loss= 0.0469, Contrastive_Depth_loss= 0.0045
epoch:161, Val: val_threshold= 0.0009, val_ACC= 0.4496, val_ACER= 0.5664
epoch:161, Test: ACC= 0.5233, APCER= 0.4265, BPCER= 0.6667, ACER= 0.5466
epoch:162, Train: Absolute_Depth_loss= 0.0470, Contrastive_Depth_loss= 0.0045
epoch:163, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:164, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:165, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:166, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:167, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:168, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0045
epoch:169, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:170, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:171, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:172, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:173, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:174, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:175, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:176, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:177, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:178, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:179, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:180, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0045
epoch:181, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:181, Val: val_threshold= 0.0008, val_ACC= 0.3953, val_ACER= 0.5943
epoch:181, Test: ACC= 0.6860, APCER= 0.2647, BPCER= 0.5000, ACER= 0.3824
epoch:182, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:183, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0045
epoch:184, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:185, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0045
epoch:186, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:187, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:188, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:189, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:190, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:191, Train: Absolute_Depth_loss= 0.0467, Contrastive_Depth_loss= 0.0045
epoch:192, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:193, Train: Absolute_Depth_loss= 0.0466, Contrastive_Depth_loss= 0.0045
epoch:194, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:195, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0044
epoch:196, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:197, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:198, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0045
epoch:199, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0045
epoch:200, Train: Absolute_Depth_loss= 0.0465, Contrastive_Depth_loss= 0.0045
epoch:201, Train: Absolute_Depth_loss= 0.0464, Contrastive_Depth_loss= 0.0045
epoch:201, Val: val_threshold= 0.0007, val_ACC= 0.5504, val_ACER= 0.4477
epoch:201, Test: ACC= 0.5233, APCER= 0.5077, BPCER= 0.3810, ACER= 0.4443
epoch:202, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0045
epoch:203, Train: Absolute_Depth_loss= 0.0462, Contrastive_Depth_loss= 0.0045
epoch:204, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0045
epoch:205, Train: Absolute_Depth_loss= 0.0461, Contrastive_Depth_loss= 0.0045
epoch:206, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:207, Train: Absolute_Depth_loss= 0.0454, Contrastive_Depth_loss= 0.0045
epoch:208, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:209, Train: Absolute_Depth_loss= 0.0456, Contrastive_Depth_loss= 0.0045
epoch:210, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0045
epoch:211, Train: Absolute_Depth_loss= 0.0458, Contrastive_Depth_loss= 0.0045
epoch:212, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:213, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0045
epoch:214, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0045
epoch:215, Train: Absolute_Depth_loss= 0.0456, Contrastive_Depth_loss= 0.0045
epoch:216, Train: Absolute_Depth_loss= 0.0460, Contrastive_Depth_loss= 0.0045
epoch:217, Train: Absolute_Depth_loss= 0.0463, Contrastive_Depth_loss= 0.0046
epoch:218, Train: Absolute_Depth_loss= 0.0453, Contrastive_Depth_loss= 0.0045
epoch:219, Train: Absolute_Depth_loss= 0.0451, Contrastive_Depth_loss= 0.0045
epoch:220, Train: Absolute_Depth_loss= 0.0456, Contrastive_Depth_loss= 0.0045
epoch:221, Train: Absolute_Depth_loss= 0.0459, Contrastive_Depth_loss= 0.0046
epoch:221, Val: val_threshold= 0.0009, val_ACC= 0.5891, val_ACER= 0.4292
epoch:221, Test: ACC= 0.3023, APCER= 0.8056, BPCER= 0.1429, ACER= 0.4742
epoch:222, Train: Absolute_Depth_loss= 0.0455, Contrastive_Depth_loss= 0.0046
epoch:223, Train: Absolute_Depth_loss= 0.0452, Contrastive_Depth_loss= 0.0045
epoch:224, Train: Absolute_Depth_loss= 0.0451, Contrastive_Depth_loss= 0.0046
epoch:225, Train: Absolute_Depth_loss= 0.0455, Contrastive_Depth_loss= 0.0046
epoch:226, Train: Absolute_Depth_loss= 0.0457, Contrastive_Depth_loss= 0.0046
epoch:227, Train: Absolute_Depth_loss= 0.0453, Contrastive_Depth_loss= 0.0046
epoch:228, Train: Absolute_Depth_loss= 0.0452, Contrastive_Depth_loss= 0.0046
epoch:229, Train: Absolute_Depth_loss= 0.0449, Contrastive_Depth_loss= 0.0046
epoch:230, Train: Absolute_Depth_loss= 0.0446, Contrastive_Depth_loss= 0.0046
epoch:231, Train: Absolute_Depth_loss= 0.0456, Contrastive_Depth_loss= 0.0047
epoch:232, Train: Absolute_Depth_loss= 0.0452, Contrastive_Depth_loss= 0.0046
epoch:233, Train: Absolute_Depth_loss= 0.0450, Contrastive_Depth_loss= 0.0046
epoch:234, Train: Absolute_Depth_loss= 0.0451, Contrastive_Depth_loss= 0.0046
epoch:235, Train: Absolute_Depth_loss= 0.0450, Contrastive_Depth_loss= 0.0047
epoch:236, Train: Absolute_Depth_loss= 0.0453, Contrastive_Depth_loss= 0.0047
epoch:237, Train: Absolute_Depth_loss= 0.0448, Contrastive_Depth_loss= 0.0047
epoch:238, Train: Absolute_Depth_loss= 0.0444, Contrastive_Depth_loss= 0.0047
epoch:239, Train: Absolute_Depth_loss= 0.0446, Contrastive_Depth_loss= 0.0047
epoch:240, Train: Absolute_Depth_loss= 0.0447, Contrastive_Depth_loss= 0.0047
epoch:241, Train: Absolute_Depth_loss= 0.0446, Contrastive_Depth_loss= 0.0047
epoch:241, Val: val_threshold= 0.0007, val_ACC= 0.5969, val_ACER= 0.4047
epoch:241, Test: ACC= 0.6047, APCER= 0.4242, BPCER= 0.3000, ACER= 0.3621
epoch:242, Train: Absolute_Depth_loss= 0.0444, Contrastive_Depth_loss= 0.0047
epoch:243, Train: Absolute_Depth_loss= 0.0443, Contrastive_Depth_loss= 0.0047
epoch:244, Train: Absolute_Depth_loss= 0.0438, Contrastive_Depth_loss= 0.0047
epoch:245, Train: Absolute_Depth_loss= 0.0440, Contrastive_Depth_loss= 0.0047
epoch:246, Train: Absolute_Depth_loss= 0.0444, Contrastive_Depth_loss= 0.0047
epoch:247, Train: Absolute_Depth_loss= 0.0450, Contrastive_Depth_loss= 0.0047
epoch:248, Train: Absolute_Depth_loss= 0.0442, Contrastive_Depth_loss= 0.0047
epoch:249, Train: Absolute_Depth_loss= 0.0444, Contrastive_Depth_loss= 0.0047
epoch:250, Train: Absolute_Depth_loss= 0.0434, Contrastive_Depth_loss= 0.0047
epoch:251, Train: Absolute_Depth_loss= 0.0441, Contrastive_Depth_loss= 0.0048
epoch:252, Train: Absolute_Depth_loss= 0.0445, Contrastive_Depth_loss= 0.0048
epoch:253, Train: Absolute_Depth_loss= 0.0439, Contrastive_Depth_loss= 0.0048
epoch:254, Train: Absolute_Depth_loss= 0.0431, Contrastive_Depth_loss= 0.0047
epoch:255, Train: Absolute_Depth_loss= 0.0433, Contrastive_Depth_loss= 0.0048
epoch:256, Train: Absolute_Depth_loss= 0.0429, Contrastive_Depth_loss= 0.0048
epoch:257, Train: Absolute_Depth_loss= 0.0437, Contrastive_Depth_loss= 0.0048
epoch:258, Train: Absolute_Depth_loss= 0.0432, Contrastive_Depth_loss= 0.0048
epoch:259, Train: Absolute_Depth_loss= 0.0437, Contrastive_Depth_loss= 0.0048
epoch:260, Train: Absolute_Depth_loss= 0.0431, Contrastive_Depth_loss= 0.0047
epoch:261, Train: Absolute_Depth_loss= 0.0424, Contrastive_Depth_loss= 0.0049
epoch:261, Val: val_threshold= 0.0005, val_ACC= 0.5426, val_ACER= 0.4660
epoch:261, Test: ACC= 0.3953, APCER= 0.6986, BPCER= 0.0769, ACER= 0.3878
epoch:262, Train: Absolute_Depth_loss= 0.0439, Contrastive_Depth_loss= 0.0048
epoch:263, Train: Absolute_Depth_loss= 0.0426, Contrastive_Depth_loss= 0.0049
epoch:264, Train: Absolute_Depth_loss= 0.0424, Contrastive_Depth_loss= 0.0049
epoch:265, Train: Absolute_Depth_loss= 0.0423, Contrastive_Depth_loss= 0.0048
epoch:266, Train: Absolute_Depth_loss= 0.0425, Contrastive_Depth_loss= 0.0049
epoch:267, Train: Absolute_Depth_loss= 0.0425, Contrastive_Depth_loss= 0.0049
epoch:268, Train: Absolute_Depth_loss= 0.0416, Contrastive_Depth_loss= 0.0049
epoch:269, Train: Absolute_Depth_loss= 0.0424, Contrastive_Depth_loss= 0.0050
epoch:270, Train: Absolute_Depth_loss= 0.0426, Contrastive_Depth_loss= 0.0049
epoch:271, Train: Absolute_Depth_loss= 0.0422, Contrastive_Depth_loss= 0.0049
epoch:272, Train: Absolute_Depth_loss= 0.0414, Contrastive_Depth_loss= 0.0049
epoch:273, Train: Absolute_Depth_loss= 0.0420, Contrastive_Depth_loss= 0.0051
epoch:274, Train: Absolute_Depth_loss= 0.0418, Contrastive_Depth_loss= 0.0051
epoch:275, Train: Absolute_Depth_loss= 0.0419, Contrastive_Depth_loss= 0.0050
epoch:276, Train: Absolute_Depth_loss= 0.0408, Contrastive_Depth_loss= 0.0050
epoch:277, Train: Absolute_Depth_loss= 0.0412, Contrastive_Depth_loss= 0.0050
epoch:278, Train: Absolute_Depth_loss= 0.0410, Contrastive_Depth_loss= 0.0050
epoch:279, Train: Absolute_Depth_loss= 0.0412, Contrastive_Depth_loss= 0.0050
epoch:280, Train: Absolute_Depth_loss= 0.0410, Contrastive_Depth_loss= 0.0050
epoch:281, Train: Absolute_Depth_loss= 0.0410, Contrastive_Depth_loss= 0.0051
epoch:281, Val: val_threshold= 0.0007, val_ACC= 0.6047, val_ACER= 0.4019
epoch:281, Test: ACC= 0.5116, APCER= 0.4861, BPCER= 0.5000, ACER= 0.4931
epoch:282, Train: Absolute_Depth_loss= 0.0409, Contrastive_Depth_loss= 0.0050
epoch:283, Train: Absolute_Depth_loss= 0.0414, Contrastive_Depth_loss= 0.0051
epoch:284, Train: Absolute_Depth_loss= 0.0414, Contrastive_Depth_loss= 0.0051
epoch:285, Train: Absolute_Depth_loss= 0.0403, Contrastive_Depth_loss= 0.0051
epoch:286, Train: Absolute_Depth_loss= 0.0404, Contrastive_Depth_loss= 0.0051
epoch:287, Train: Absolute_Depth_loss= 0.0405, Contrastive_Depth_loss= 0.0051
epoch:288, Train: Absolute_Depth_loss= 0.0402, Contrastive_Depth_loss= 0.0050
epoch:289, Train: Absolute_Depth_loss= 0.0403, Contrastive_Depth_loss= 0.0052
epoch:290, Train: Absolute_Depth_loss= 0.0406, Contrastive_Depth_loss= 0.0051
epoch:291, Train: Absolute_Depth_loss= 0.0396, Contrastive_Depth_loss= 0.0052
epoch:292, Train: Absolute_Depth_loss= 0.0390, Contrastive_Depth_loss= 0.0052
epoch:293, Train: Absolute_Depth_loss= 0.0402, Contrastive_Depth_loss= 0.0051
epoch:294, Train: Absolute_Depth_loss= 0.0405, Contrastive_Depth_loss= 0.0051
epoch:295, Train: Absolute_Depth_loss= 0.0391, Contrastive_Depth_loss= 0.0052
epoch:296, Train: Absolute_Depth_loss= 0.0399, Contrastive_Depth_loss= 0.0052
epoch:297, Train: Absolute_Depth_loss= 0.0383, Contrastive_Depth_loss= 0.0052
epoch:298, Train: Absolute_Depth_loss= 0.0385, Contrastive_Depth_loss= 0.0052
epoch:299, Train: Absolute_Depth_loss= 0.0387, Contrastive_Depth_loss= 0.0052
epoch:300, Train: Absolute_Depth_loss= 0.0395, Contrastive_Depth_loss= 0.0052
epoch:301, Train: Absolute_Depth_loss= 0.0391, Contrastive_Depth_loss= 0.0052
epoch:301, Val: val_threshold= 0.0008, val_ACC= 0.5659, val_ACER= 0.4401
epoch:301, Test: ACC= 0.3837, APCER= 0.7833, BPCER= 0.2308, ACER= 0.5071
epoch:302, Train: Absolute_Depth_loss= 0.0394, Contrastive_Depth_loss= 0.0052
epoch:303, Train: Absolute_Depth_loss= 0.0384, Contrastive_Depth_loss= 0.0052
epoch:304, Train: Absolute_Depth_loss= 0.0393, Contrastive_Depth_loss= 0.0052
epoch:305, Train: Absolute_Depth_loss= 0.0382, Contrastive_Depth_loss= 0.0053
epoch:306, Train: Absolute_Depth_loss= 0.0390, Contrastive_Depth_loss= 0.0053
epoch:307, Train: Absolute_Depth_loss= 0.0380, Contrastive_Depth_loss= 0.0053
epoch:308, Train: Absolute_Depth_loss= 0.0385, Contrastive_Depth_loss= 0.0052
epoch:309, Train: Absolute_Depth_loss= 0.0374, Contrastive_Depth_loss= 0.0054
epoch:310, Train: Absolute_Depth_loss= 0.0376, Contrastive_Depth_loss= 0.0053
epoch:311, Train: Absolute_Depth_loss= 0.0379, Contrastive_Depth_loss= 0.0053
epoch:312, Train: Absolute_Depth_loss= 0.0379, Contrastive_Depth_loss= 0.0053
epoch:313, Train: Absolute_Depth_loss= 0.0369, Contrastive_Depth_loss= 0.0053
epoch:314, Train: Absolute_Depth_loss= 0.0375, Contrastive_Depth_loss= 0.0053
epoch:315, Train: Absolute_Depth_loss= 0.0368, Contrastive_Depth_loss= 0.0054
epoch:316, Train: Absolute_Depth_loss= 0.0373, Contrastive_Depth_loss= 0.0054
epoch:317, Train: Absolute_Depth_loss= 0.0370, Contrastive_Depth_loss= 0.0054
epoch:318, Train: Absolute_Depth_loss= 0.0356, Contrastive_Depth_loss= 0.0053
epoch:319, Train: Absolute_Depth_loss= 0.0365, Contrastive_Depth_loss= 0.0054
epoch:320, Train: Absolute_Depth_loss= 0.0365, Contrastive_Depth_loss= 0.0054
epoch:321, Train: Absolute_Depth_loss= 0.0371, Contrastive_Depth_loss= 0.0053
epoch:321, Val: val_threshold= 0.0009, val_ACC= 0.6589, val_ACER= 0.3960
epoch:321, Test: ACC= 0.5349, APCER= 0.3881, BPCER= 0.7368, ACER= 0.5625
epoch:322, Train: Absolute_Depth_loss= 0.0359, Contrastive_Depth_loss= 0.0054
epoch:323, Train: Absolute_Depth_loss= 0.0369, Contrastive_Depth_loss= 0.0054
epoch:324, Train: Absolute_Depth_loss= 0.0363, Contrastive_Depth_loss= 0.0054
epoch:325, Train: Absolute_Depth_loss= 0.0360, Contrastive_Depth_loss= 0.0054
epoch:326, Train: Absolute_Depth_loss= 0.0361, Contrastive_Depth_loss= 0.0054
epoch:327, Train: Absolute_Depth_loss= 0.0353, Contrastive_Depth_loss= 0.0055
epoch:328, Train: Absolute_Depth_loss= 0.0361, Contrastive_Depth_loss= 0.0055
epoch:329, Train: Absolute_Depth_loss= 0.0353, Contrastive_Depth_loss= 0.0054
epoch:330, Train: Absolute_Depth_loss= 0.0345, Contrastive_Depth_loss= 0.0055
epoch:331, Train: Absolute_Depth_loss= 0.0352, Contrastive_Depth_loss= 0.0055
epoch:332, Train: Absolute_Depth_loss= 0.0349, Contrastive_Depth_loss= 0.0055
epoch:333, Train: Absolute_Depth_loss= 0.0341, Contrastive_Depth_loss= 0.0056
epoch:334, Train: Absolute_Depth_loss= 0.0347, Contrastive_Depth_loss= 0.0055
epoch:335, Train: Absolute_Depth_loss= 0.0344, Contrastive_Depth_loss= 0.0056
epoch:336, Train: Absolute_Depth_loss= 0.0345, Contrastive_Depth_loss= 0.0056
epoch:337, Train: Absolute_Depth_loss= 0.0345, Contrastive_Depth_loss= 0.0055
epoch:338, Train: Absolute_Depth_loss= 0.0335, Contrastive_Depth_loss= 0.0055
epoch:339, Train: Absolute_Depth_loss= 0.0340, Contrastive_Depth_loss= 0.0055
epoch:340, Train: Absolute_Depth_loss= 0.0341, Contrastive_Depth_loss= 0.0055
epoch:341, Train: Absolute_Depth_loss= 0.0331, Contrastive_Depth_loss= 0.0055
epoch:341, Val: val_threshold= 0.0007, val_ACC= 0.5891, val_ACER= 0.4202
epoch:341, Test: ACC= 0.4884, APCER= 0.5152, BPCER= 0.5000, ACER= 0.5076
epoch:342, Train: Absolute_Depth_loss= 0.0336, Contrastive_Depth_loss= 0.0056
epoch:343, Train: Absolute_Depth_loss= 0.0338, Contrastive_Depth_loss= 0.0056
epoch:344, Train: Absolute_Depth_loss= 0.0337, Contrastive_Depth_loss= 0.0055
epoch:345, Train: Absolute_Depth_loss= 0.0335, Contrastive_Depth_loss= 0.0056
epoch:346, Train: Absolute_Depth_loss= 0.0331, Contrastive_Depth_loss= 0.0056
epoch:347, Train: Absolute_Depth_loss= 0.0334, Contrastive_Depth_loss= 0.0056
epoch:348, Train: Absolute_Depth_loss= 0.0332, Contrastive_Depth_loss= 0.0056
epoch:349, Train: Absolute_Depth_loss= 0.0326, Contrastive_Depth_loss= 0.0056
epoch:350, Train: Absolute_Depth_loss= 0.0332, Contrastive_Depth_loss= 0.0056
epoch:351, Train: Absolute_Depth_loss= 0.0334, Contrastive_Depth_loss= 0.0056
epoch:352, Train: Absolute_Depth_loss= 0.0317, Contrastive_Depth_loss= 0.0056
epoch:353, Train: Absolute_Depth_loss= 0.0322, Contrastive_Depth_loss= 0.0057
epoch:354, Train: Absolute_Depth_loss= 0.0319, Contrastive_Depth_loss= 0.0056
epoch:355, Train: Absolute_Depth_loss= 0.0317, Contrastive_Depth_loss= 0.0056
epoch:356, Train: Absolute_Depth_loss= 0.0324, Contrastive_Depth_loss= 0.0056
epoch:357, Train: Absolute_Depth_loss= 0.0323, Contrastive_Depth_loss= 0.0057
epoch:358, Train: Absolute_Depth_loss= 0.0322, Contrastive_Depth_loss= 0.0056
epoch:359, Train: Absolute_Depth_loss= 0.0320, Contrastive_Depth_loss= 0.0057
epoch:360, Train: Absolute_Depth_loss= 0.0309, Contrastive_Depth_loss= 0.0056
epoch:361, Train: Absolute_Depth_loss= 0.0308, Contrastive_Depth_loss= 0.0057
epoch:361, Val: val_threshold= 0.0010, val_ACC= 0.5116, val_ACER= 0.5026
epoch:361, Test: ACC= 0.4186, APCER= 0.6721, BPCER= 0.3600, ACER= 0.5161
epoch:362, Train: Absolute_Depth_loss= 0.0323, Contrastive_Depth_loss= 0.0057
epoch:363, Train: Absolute_Depth_loss= 0.0310, Contrastive_Depth_loss= 0.0057
epoch:364, Train: Absolute_Depth_loss= 0.0307, Contrastive_Depth_loss= 0.0057
epoch:365, Train: Absolute_Depth_loss= 0.0324, Contrastive_Depth_loss= 0.0057
epoch:366, Train: Absolute_Depth_loss= 0.0307, Contrastive_Depth_loss= 0.0057
epoch:367, Train: Absolute_Depth_loss= 0.0322, Contrastive_Depth_loss= 0.0057
epoch:368, Train: Absolute_Depth_loss= 0.0304, Contrastive_Depth_loss= 0.0056
epoch:369, Train: Absolute_Depth_loss= 0.0303, Contrastive_Depth_loss= 0.0057
epoch:370, Train: Absolute_Depth_loss= 0.0305, Contrastive_Depth_loss= 0.0057
epoch:371, Train: Absolute_Depth_loss= 0.0307, Contrastive_Depth_loss= 0.0057
epoch:372, Train: Absolute_Depth_loss= 0.0303, Contrastive_Depth_loss= 0.0057
epoch:373, Train: Absolute_Depth_loss= 0.0286, Contrastive_Depth_loss= 0.0057
epoch:374, Train: Absolute_Depth_loss= 0.0298, Contrastive_Depth_loss= 0.0057
epoch:375, Train: Absolute_Depth_loss= 0.0303, Contrastive_Depth_loss= 0.0057
epoch:376, Train: Absolute_Depth_loss= 0.0290, Contrastive_Depth_loss= 0.0057
epoch:377, Train: Absolute_Depth_loss= 0.0290, Contrastive_Depth_loss= 0.0057
epoch:378, Train: Absolute_Depth_loss= 0.0296, Contrastive_Depth_loss= 0.0057
epoch:379, Train: Absolute_Depth_loss= 0.0296, Contrastive_Depth_loss= 0.0057
epoch:380, Train: Absolute_Depth_loss= 0.0299, Contrastive_Depth_loss= 0.0058
epoch:381, Train: Absolute_Depth_loss= 0.0313, Contrastive_Depth_loss= 0.0057
epoch:381, Val: val_threshold= 0.0010, val_ACC= 0.6047, val_ACER= 0.4141
epoch:381, Test: ACC= 0.5000, APCER= 0.4857, BPCER= 0.5625, ACER= 0.5241
epoch:382, Train: Absolute_Depth_loss= 0.0297, Contrastive_Depth_loss= 0.0057
epoch:383, Train: Absolute_Depth_loss= 0.0293, Contrastive_Depth_loss= 0.0057
epoch:384, Train: Absolute_Depth_loss= 0.0299, Contrastive_Depth_loss= 0.0058
epoch:385, Train: Absolute_Depth_loss= 0.0305, Contrastive_Depth_loss= 0.0057
epoch:386, Train: Absolute_Depth_loss= 0.0284, Contrastive_Depth_loss= 0.0058
epoch:387, Train: Absolute_Depth_loss= 0.0295, Contrastive_Depth_loss= 0.0058
epoch:388, Train: Absolute_Depth_loss= 0.0293, Contrastive_Depth_loss= 0.0058
epoch:389, Train: Absolute_Depth_loss= 0.0288, Contrastive_Depth_loss= 0.0057
epoch:390, Train: Absolute_Depth_loss= 0.0293, Contrastive_Depth_loss= 0.0057
epoch:391, Train: Absolute_Depth_loss= 0.0276, Contrastive_Depth_loss= 0.0058
epoch:392, Train: Absolute_Depth_loss= 0.0293, Contrastive_Depth_loss= 0.0058
epoch:393, Train: Absolute_Depth_loss= 0.0289, Contrastive_Depth_loss= 0.0057
epoch:394, Train: Absolute_Depth_loss= 0.0275, Contrastive_Depth_loss= 0.0058
epoch:395, Train: Absolute_Depth_loss= 0.0284, Contrastive_Depth_loss= 0.0058
epoch:396, Train: Absolute_Depth_loss= 0.0285, Contrastive_Depth_loss= 0.0058
epoch:397, Train: Absolute_Depth_loss= 0.0289, Contrastive_Depth_loss= 0.0058
epoch:398, Train: Absolute_Depth_loss= 0.0289, Contrastive_Depth_loss= 0.0058
epoch:399, Train: Absolute_Depth_loss= 0.0276, Contrastive_Depth_loss= 0.0057
epoch:400, Train: Absolute_Depth_loss= 0.0281, Contrastive_Depth_loss= 0.0057
epoch:401, Train: Absolute_Depth_loss= 0.0270, Contrastive_Depth_loss= 0.0057
epoch:401, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4879
epoch:401, Test: ACC= 0.3256, APCER= 0.7612, BPCER= 0.3684, ACER= 0.5648
epoch:402, Train: Absolute_Depth_loss= 0.0273, Contrastive_Depth_loss= 0.0057
epoch:403, Train: Absolute_Depth_loss= 0.0281, Contrastive_Depth_loss= 0.0058
epoch:404, Train: Absolute_Depth_loss= 0.0274, Contrastive_Depth_loss= 0.0057
epoch:405, Train: Absolute_Depth_loss= 0.0277, Contrastive_Depth_loss= 0.0057
epoch:406, Train: Absolute_Depth_loss= 0.0273, Contrastive_Depth_loss= 0.0058
epoch:407, Train: Absolute_Depth_loss= 0.0275, Contrastive_Depth_loss= 0.0057
epoch:408, Train: Absolute_Depth_loss= 0.0274, Contrastive_Depth_loss= 0.0058
epoch:409, Train: Absolute_Depth_loss= 0.0258, Contrastive_Depth_loss= 0.0058
epoch:410, Train: Absolute_Depth_loss= 0.0271, Contrastive_Depth_loss= 0.0058
epoch:411, Train: Absolute_Depth_loss= 0.0272, Contrastive_Depth_loss= 0.0058
epoch:412, Train: Absolute_Depth_loss= 0.0261, Contrastive_Depth_loss= 0.0057
epoch:413, Train: Absolute_Depth_loss= 0.0278, Contrastive_Depth_loss= 0.0058
epoch:414, Train: Absolute_Depth_loss= 0.0255, Contrastive_Depth_loss= 0.0057
epoch:415, Train: Absolute_Depth_loss= 0.0264, Contrastive_Depth_loss= 0.0058
epoch:416, Train: Absolute_Depth_loss= 0.0268, Contrastive_Depth_loss= 0.0059
epoch:417, Train: Absolute_Depth_loss= 0.0259, Contrastive_Depth_loss= 0.0058
epoch:418, Train: Absolute_Depth_loss= 0.0261, Contrastive_Depth_loss= 0.0058
epoch:419, Train: Absolute_Depth_loss= 0.0262, Contrastive_Depth_loss= 0.0058
epoch:420, Train: Absolute_Depth_loss= 0.0253, Contrastive_Depth_loss= 0.0058
epoch:421, Train: Absolute_Depth_loss= 0.0265, Contrastive_Depth_loss= 0.0058
epoch:421, Val: val_threshold= 0.0006, val_ACC= 0.5581, val_ACER= 0.4469
epoch:421, Test: ACC= 0.5233, APCER= 0.5507, BPCER= 0.1765, ACER= 0.3636
epoch:422, Train: Absolute_Depth_loss= 0.0261, Contrastive_Depth_loss= 0.0058
epoch:423, Train: Absolute_Depth_loss= 0.0265, Contrastive_Depth_loss= 0.0058
epoch:424, Train: Absolute_Depth_loss= 0.0259, Contrastive_Depth_loss= 0.0058
epoch:425, Train: Absolute_Depth_loss= 0.0267, Contrastive_Depth_loss= 0.0058
epoch:426, Train: Absolute_Depth_loss= 0.0248, Contrastive_Depth_loss= 0.0057
epoch:427, Train: Absolute_Depth_loss= 0.0260, Contrastive_Depth_loss= 0.0058
epoch:428, Train: Absolute_Depth_loss= 0.0251, Contrastive_Depth_loss= 0.0056
epoch:429, Train: Absolute_Depth_loss= 0.0252, Contrastive_Depth_loss= 0.0058
epoch:430, Train: Absolute_Depth_loss= 0.0258, Contrastive_Depth_loss= 0.0058
epoch:431, Train: Absolute_Depth_loss= 0.0258, Contrastive_Depth_loss= 0.0058
epoch:432, Train: Absolute_Depth_loss= 0.0247, Contrastive_Depth_loss= 0.0057
epoch:433, Train: Absolute_Depth_loss= 0.0264, Contrastive_Depth_loss= 0.0058
epoch:434, Train: Absolute_Depth_loss= 0.0254, Contrastive_Depth_loss= 0.0058
epoch:435, Train: Absolute_Depth_loss= 0.0253, Contrastive_Depth_loss= 0.0058
epoch:436, Train: Absolute_Depth_loss= 0.0248, Contrastive_Depth_loss= 0.0058
epoch:437, Train: Absolute_Depth_loss= 0.0257, Contrastive_Depth_loss= 0.0058
epoch:438, Train: Absolute_Depth_loss= 0.0253, Contrastive_Depth_loss= 0.0058
epoch:439, Train: Absolute_Depth_loss= 0.0249, Contrastive_Depth_loss= 0.0058
epoch:440, Train: Absolute_Depth_loss= 0.0243, Contrastive_Depth_loss= 0.0058
epoch:441, Train: Absolute_Depth_loss= 0.0242, Contrastive_Depth_loss= 0.0058
epoch:441, Val: val_threshold= 0.0006, val_ACC= 0.4884, val_ACER= 0.5074
epoch:441, Test: ACC= 0.5465, APCER= 0.4667, BPCER= 0.3636, ACER= 0.4152
epoch:442, Train: Absolute_Depth_loss= 0.0238, Contrastive_Depth_loss= 0.0057
epoch:443, Train: Absolute_Depth_loss= 0.0246, Contrastive_Depth_loss= 0.0058
epoch:444, Train: Absolute_Depth_loss= 0.0241, Contrastive_Depth_loss= 0.0058
epoch:445, Train: Absolute_Depth_loss= 0.0241, Contrastive_Depth_loss= 0.0058
epoch:446, Train: Absolute_Depth_loss= 0.0242, Contrastive_Depth_loss= 0.0058
epoch:447, Train: Absolute_Depth_loss= 0.0239, Contrastive_Depth_loss= 0.0057
epoch:448, Train: Absolute_Depth_loss= 0.0242, Contrastive_Depth_loss= 0.0058
epoch:449, Train: Absolute_Depth_loss= 0.0238, Contrastive_Depth_loss= 0.0058
epoch:450, Train: Absolute_Depth_loss= 0.0240, Contrastive_Depth_loss= 0.0058
epoch:451, Train: Absolute_Depth_loss= 0.0240, Contrastive_Depth_loss= 0.0058
epoch:452, Train: Absolute_Depth_loss= 0.0232, Contrastive_Depth_loss= 0.0057
epoch:453, Train: Absolute_Depth_loss= 0.0241, Contrastive_Depth_loss= 0.0058
epoch:454, Train: Absolute_Depth_loss= 0.0230, Contrastive_Depth_loss= 0.0057
epoch:455, Train: Absolute_Depth_loss= 0.0244, Contrastive_Depth_loss= 0.0058
epoch:456, Train: Absolute_Depth_loss= 0.0232, Contrastive_Depth_loss= 0.0057
epoch:457, Train: Absolute_Depth_loss= 0.0241, Contrastive_Depth_loss= 0.0058
epoch:458, Train: Absolute_Depth_loss= 0.0235, Contrastive_Depth_loss= 0.0058
epoch:459, Train: Absolute_Depth_loss= 0.0229, Contrastive_Depth_loss= 0.0058
epoch:460, Train: Absolute_Depth_loss= 0.0242, Contrastive_Depth_loss= 0.0058
epoch:461, Train: Absolute_Depth_loss= 0.0232, Contrastive_Depth_loss= 0.0058
epoch:461, Val: val_threshold= 0.0005, val_ACC= 0.5659, val_ACER= 0.4511
epoch:461, Test: ACC= 0.4070, APCER= 0.5970, BPCER= 0.5789, ACER= 0.5880
epoch:462, Train: Absolute_Depth_loss= 0.0238, Contrastive_Depth_loss= 0.0058
epoch:463, Train: Absolute_Depth_loss= 0.0242, Contrastive_Depth_loss= 0.0058
epoch:464, Train: Absolute_Depth_loss= 0.0235, Contrastive_Depth_loss= 0.0057
epoch:465, Train: Absolute_Depth_loss= 0.0234, Contrastive_Depth_loss= 0.0057
epoch:466, Train: Absolute_Depth_loss= 0.0229, Contrastive_Depth_loss= 0.0057
epoch:467, Train: Absolute_Depth_loss= 0.0232, Contrastive_Depth_loss= 0.0057
epoch:468, Train: Absolute_Depth_loss= 0.0229, Contrastive_Depth_loss= 0.0058
epoch:469, Train: Absolute_Depth_loss= 0.0225, Contrastive_Depth_loss= 0.0058
epoch:470, Train: Absolute_Depth_loss= 0.0228, Contrastive_Depth_loss= 0.0057
epoch:471, Train: Absolute_Depth_loss= 0.0226, Contrastive_Depth_loss= 0.0058
epoch:472, Train: Absolute_Depth_loss= 0.0236, Contrastive_Depth_loss= 0.0058
epoch:473, Train: Absolute_Depth_loss= 0.0230, Contrastive_Depth_loss= 0.0058
epoch:474, Train: Absolute_Depth_loss= 0.0229, Contrastive_Depth_loss= 0.0057
epoch:475, Train: Absolute_Depth_loss= 0.0228, Contrastive_Depth_loss= 0.0057
epoch:476, Train: Absolute_Depth_loss= 0.0227, Contrastive_Depth_loss= 0.0057
epoch:477, Train: Absolute_Depth_loss= 0.0219, Contrastive_Depth_loss= 0.0057
epoch:478, Train: Absolute_Depth_loss= 0.0217, Contrastive_Depth_loss= 0.0057
epoch:479, Train: Absolute_Depth_loss= 0.0224, Contrastive_Depth_loss= 0.0058
epoch:480, Train: Absolute_Depth_loss= 0.0225, Contrastive_Depth_loss= 0.0057
epoch:481, Train: Absolute_Depth_loss= 0.0217, Contrastive_Depth_loss= 0.0057
epoch:481, Val: val_threshold= 0.0004, val_ACC= 0.4729, val_ACER= 0.5331
epoch:481, Test: ACC= 0.5930, APCER= 0.4507, BPCER= 0.2000, ACER= 0.3254
epoch:482, Train: Absolute_Depth_loss= 0.0223, Contrastive_Depth_loss= 0.0057
epoch:483, Train: Absolute_Depth_loss= 0.0228, Contrastive_Depth_loss= 0.0058
epoch:484, Train: Absolute_Depth_loss= 0.0237, Contrastive_Depth_loss= 0.0058
epoch:485, Train: Absolute_Depth_loss= 0.0220, Contrastive_Depth_loss= 0.0057
epoch:486, Train: Absolute_Depth_loss= 0.0221, Contrastive_Depth_loss= 0.0057
epoch:487, Train: Absolute_Depth_loss= 0.0213, Contrastive_Depth_loss= 0.0057
epoch:488, Train: Absolute_Depth_loss= 0.0212, Contrastive_Depth_loss= 0.0057
epoch:489, Train: Absolute_Depth_loss= 0.0213, Contrastive_Depth_loss= 0.0057
epoch:490, Train: Absolute_Depth_loss= 0.0235, Contrastive_Depth_loss= 0.0058
epoch:491, Train: Absolute_Depth_loss= 0.0216, Contrastive_Depth_loss= 0.0057
epoch:492, Train: Absolute_Depth_loss= 0.0209, Contrastive_Depth_loss= 0.0057
epoch:493, Train: Absolute_Depth_loss= 0.0211, Contrastive_Depth_loss= 0.0056
epoch:494, Train: Absolute_Depth_loss= 0.0215, Contrastive_Depth_loss= 0.0057
epoch:495, Train: Absolute_Depth_loss= 0.0210, Contrastive_Depth_loss= 0.0057
epoch:496, Train: Absolute_Depth_loss= 0.0227, Contrastive_Depth_loss= 0.0058
epoch:497, Train: Absolute_Depth_loss= 0.0209, Contrastive_Depth_loss= 0.0057
epoch:498, Train: Absolute_Depth_loss= 0.0213, Contrastive_Depth_loss= 0.0058
epoch:499, Train: Absolute_Depth_loss= 0.0222, Contrastive_Depth_loss= 0.0057
epoch:500, Train: Absolute_Depth_loss= 0.0195, Contrastive_Depth_loss= 0.0056
epoch:501, Train: Absolute_Depth_loss= 0.0188, Contrastive_Depth_loss= 0.0055
epoch:501, Val: val_threshold= 0.0008, val_ACC= 0.6202, val_ACER= 0.3919
epoch:501, Test: ACC= 0.5233, APCER= 0.4366, BPCER= 0.6667, ACER= 0.5516
epoch:502, Train: Absolute_Depth_loss= 0.0186, Contrastive_Depth_loss= 0.0056
epoch:503, Train: Absolute_Depth_loss= 0.0180, Contrastive_Depth_loss= 0.0055
epoch:504, Train: Absolute_Depth_loss= 0.0187, Contrastive_Depth_loss= 0.0055
epoch:505, Train: Absolute_Depth_loss= 0.0182, Contrastive_Depth_loss= 0.0056
epoch:506, Train: Absolute_Depth_loss= 0.0184, Contrastive_Depth_loss= 0.0055
epoch:507, Train: Absolute_Depth_loss= 0.0176, Contrastive_Depth_loss= 0.0055
epoch:508, Train: Absolute_Depth_loss= 0.0181, Contrastive_Depth_loss= 0.0055
epoch:509, Train: Absolute_Depth_loss= 0.0176, Contrastive_Depth_loss= 0.0054
epoch:510, Train: Absolute_Depth_loss= 0.0180, Contrastive_Depth_loss= 0.0055
epoch:511, Train: Absolute_Depth_loss= 0.0177, Contrastive_Depth_loss= 0.0055
epoch:512, Train: Absolute_Depth_loss= 0.0177, Contrastive_Depth_loss= 0.0055
epoch:513, Train: Absolute_Depth_loss= 0.0177, Contrastive_Depth_loss= 0.0055
epoch:514, Train: Absolute_Depth_loss= 0.0176, Contrastive_Depth_loss= 0.0055
epoch:515, Train: Absolute_Depth_loss= 0.0170, Contrastive_Depth_loss= 0.0055
epoch:516, Train: Absolute_Depth_loss= 0.0173, Contrastive_Depth_loss= 0.0055
epoch:517, Train: Absolute_Depth_loss= 0.0173, Contrastive_Depth_loss= 0.0055
epoch:518, Train: Absolute_Depth_loss= 0.0180, Contrastive_Depth_loss= 0.0056
epoch:519, Train: Absolute_Depth_loss= 0.0170, Contrastive_Depth_loss= 0.0054
epoch:520, Train: Absolute_Depth_loss= 0.0169, Contrastive_Depth_loss= 0.0054
epoch:521, Train: Absolute_Depth_loss= 0.0170, Contrastive_Depth_loss= 0.0055
epoch:521, Val: val_threshold= 0.0006, val_ACC= 0.5194, val_ACER= 0.4804
epoch:521, Test: ACC= 0.5116, APCER= 0.4783, BPCER= 0.5294, ACER= 0.5038
epoch:522, Train: Absolute_Depth_loss= 0.0166, Contrastive_Depth_loss= 0.0055
epoch:523, Train: Absolute_Depth_loss= 0.0163, Contrastive_Depth_loss= 0.0054
epoch:524, Train: Absolute_Depth_loss= 0.0161, Contrastive_Depth_loss= 0.0054
epoch:525, Train: Absolute_Depth_loss= 0.0163, Contrastive_Depth_loss= 0.0053
epoch:526, Train: Absolute_Depth_loss= 0.0168, Contrastive_Depth_loss= 0.0054
epoch:527, Train: Absolute_Depth_loss= 0.0162, Contrastive_Depth_loss= 0.0054
epoch:528, Train: Absolute_Depth_loss= 0.0170, Contrastive_Depth_loss= 0.0055
epoch:529, Train: Absolute_Depth_loss= 0.0162, Contrastive_Depth_loss= 0.0055
epoch:530, Train: Absolute_Depth_loss= 0.0168, Contrastive_Depth_loss= 0.0055
epoch:531, Train: Absolute_Depth_loss= 0.0174, Contrastive_Depth_loss= 0.0055
epoch:532, Train: Absolute_Depth_loss= 0.0166, Contrastive_Depth_loss= 0.0054
epoch:533, Train: Absolute_Depth_loss= 0.0175, Contrastive_Depth_loss= 0.0055
epoch:534, Train: Absolute_Depth_loss= 0.0165, Contrastive_Depth_loss= 0.0055
epoch:535, Train: Absolute_Depth_loss= 0.0169, Contrastive_Depth_loss= 0.0055
epoch:536, Train: Absolute_Depth_loss= 0.0163, Contrastive_Depth_loss= 0.0054
epoch:537, Train: Absolute_Depth_loss= 0.0161, Contrastive_Depth_loss= 0.0054
epoch:538, Train: Absolute_Depth_loss= 0.0153, Contrastive_Depth_loss= 0.0053
epoch:539, Train: Absolute_Depth_loss= 0.0161, Contrastive_Depth_loss= 0.0054
epoch:540, Train: Absolute_Depth_loss= 0.0161, Contrastive_Depth_loss= 0.0054
epoch:541, Train: Absolute_Depth_loss= 0.0160, Contrastive_Depth_loss= 0.0053
epoch:541, Val: val_threshold= 0.0005, val_ACC= 0.4884, val_ACER= 0.5132
epoch:541, Test: ACC= 0.4419, APCER= 0.6119, BPCER= 0.3684, ACER= 0.4902
epoch:542, Train: Absolute_Depth_loss= 0.0158, Contrastive_Depth_loss= 0.0054
epoch:543, Train: Absolute_Depth_loss= 0.0165, Contrastive_Depth_loss= 0.0055
epoch:544, Train: Absolute_Depth_loss= 0.0165, Contrastive_Depth_loss= 0.0054
epoch:545, Train: Absolute_Depth_loss= 0.0167, Contrastive_Depth_loss= 0.0055
epoch:546, Train: Absolute_Depth_loss= 0.0159, Contrastive_Depth_loss= 0.0054
epoch:547, Train: Absolute_Depth_loss= 0.0156, Contrastive_Depth_loss= 0.0054
epoch:548, Train: Absolute_Depth_loss= 0.0158, Contrastive_Depth_loss= 0.0054
epoch:549, Train: Absolute_Depth_loss= 0.0149, Contrastive_Depth_loss= 0.0053
epoch:550, Train: Absolute_Depth_loss= 0.0158, Contrastive_Depth_loss= 0.0054
epoch:551, Train: Absolute_Depth_loss= 0.0154, Contrastive_Depth_loss= 0.0054
epoch:552, Train: Absolute_Depth_loss= 0.0167, Contrastive_Depth_loss= 0.0055
epoch:553, Train: Absolute_Depth_loss= 0.0156, Contrastive_Depth_loss= 0.0054
epoch:554, Train: Absolute_Depth_loss= 0.0152, Contrastive_Depth_loss= 0.0053
epoch:555, Train: Absolute_Depth_loss= 0.0153, Contrastive_Depth_loss= 0.0054
epoch:556, Train: Absolute_Depth_loss= 0.0162, Contrastive_Depth_loss= 0.0054
epoch:557, Train: Absolute_Depth_loss= 0.0150, Contrastive_Depth_loss= 0.0052
epoch:558, Train: Absolute_Depth_loss= 0.0158, Contrastive_Depth_loss= 0.0054
epoch:559, Train: Absolute_Depth_loss= 0.0150, Contrastive_Depth_loss= 0.0052
epoch:560, Train: Absolute_Depth_loss= 0.0159, Contrastive_Depth_loss= 0.0055
epoch:561, Train: Absolute_Depth_loss= 0.0151, Contrastive_Depth_loss= 0.0053
epoch:561, Val: val_threshold= 0.0005, val_ACC= 0.5271, val_ACER= 0.4886
epoch:561, Test: ACC= 0.5465, APCER= 0.4225, BPCER= 0.6000, ACER= 0.5113
epoch:562, Train: Absolute_Depth_loss= 0.0158, Contrastive_Depth_loss= 0.0054
epoch:563, Train: Absolute_Depth_loss= 0.0145, Contrastive_Depth_loss= 0.0053
epoch:564, Train: Absolute_Depth_loss= 0.0157, Contrastive_Depth_loss= 0.0054
epoch:565, Train: Absolute_Depth_loss= 0.0149, Contrastive_Depth_loss= 0.0053
epoch:566, Train: Absolute_Depth_loss= 0.0154, Contrastive_Depth_loss= 0.0053
epoch:567, Train: Absolute_Depth_loss= 0.0155, Contrastive_Depth_loss= 0.0054
epoch:568, Train: Absolute_Depth_loss= 0.0148, Contrastive_Depth_loss= 0.0053
epoch:569, Train: Absolute_Depth_loss= 0.0152, Contrastive_Depth_loss= 0.0053
epoch:570, Train: Absolute_Depth_loss= 0.0145, Contrastive_Depth_loss= 0.0053
epoch:571, Train: Absolute_Depth_loss= 0.0162, Contrastive_Depth_loss= 0.0054
epoch:572, Train: Absolute_Depth_loss= 0.0157, Contrastive_Depth_loss= 0.0054
epoch:573, Train: Absolute_Depth_loss= 0.0148, Contrastive_Depth_loss= 0.0053
epoch:574, Train: Absolute_Depth_loss= 0.0154, Contrastive_Depth_loss= 0.0053
epoch:575, Train: Absolute_Depth_loss= 0.0150, Contrastive_Depth_loss= 0.0054
epoch:576, Train: Absolute_Depth_loss= 0.0144, Contrastive_Depth_loss= 0.0053
epoch:577, Train: Absolute_Depth_loss= 0.0156, Contrastive_Depth_loss= 0.0053
epoch:578, Train: Absolute_Depth_loss= 0.0145, Contrastive_Depth_loss= 0.0052
epoch:579, Train: Absolute_Depth_loss= 0.0152, Contrastive_Depth_loss= 0.0054
epoch:580, Train: Absolute_Depth_loss= 0.0146, Contrastive_Depth_loss= 0.0053
epoch:581, Train: Absolute_Depth_loss= 0.0152, Contrastive_Depth_loss= 0.0054
epoch:581, Val: val_threshold= 0.0005, val_ACC= 0.5116, val_ACER= 0.5014
epoch:581, Test: ACC= 0.5000, APCER= 0.4839, BPCER= 0.5417, ACER= 0.5128
epoch:582, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:583, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:584, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:585, Train: Absolute_Depth_loss= 0.0146, Contrastive_Depth_loss= 0.0052
epoch:586, Train: Absolute_Depth_loss= 0.0144, Contrastive_Depth_loss= 0.0052
epoch:587, Train: Absolute_Depth_loss= 0.0143, Contrastive_Depth_loss= 0.0052
epoch:588, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:589, Train: Absolute_Depth_loss= 0.0139, Contrastive_Depth_loss= 0.0052
epoch:590, Train: Absolute_Depth_loss= 0.0137, Contrastive_Depth_loss= 0.0052
epoch:591, Train: Absolute_Depth_loss= 0.0145, Contrastive_Depth_loss= 0.0053
epoch:592, Train: Absolute_Depth_loss= 0.0143, Contrastive_Depth_loss= 0.0053
epoch:593, Train: Absolute_Depth_loss= 0.0140, Contrastive_Depth_loss= 0.0052
epoch:594, Train: Absolute_Depth_loss= 0.0149, Contrastive_Depth_loss= 0.0053
epoch:595, Train: Absolute_Depth_loss= 0.0145, Contrastive_Depth_loss= 0.0053
epoch:596, Train: Absolute_Depth_loss= 0.0142, Contrastive_Depth_loss= 0.0053
epoch:597, Train: Absolute_Depth_loss= 0.0146, Contrastive_Depth_loss= 0.0053
epoch:598, Train: Absolute_Depth_loss= 0.0149, Contrastive_Depth_loss= 0.0053
epoch:599, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:600, Train: Absolute_Depth_loss= 0.0132, Contrastive_Depth_loss= 0.0052
epoch:601, Train: Absolute_Depth_loss= 0.0146, Contrastive_Depth_loss= 0.0053
epoch:601, Val: val_threshold= 0.0007, val_ACC= 0.5039, val_ACER= 0.5086
epoch:601, Test: ACC= 0.5233, APCER= 0.5217, BPCER= 0.2941, ACER= 0.4079
epoch:602, Train: Absolute_Depth_loss= 0.0138, Contrastive_Depth_loss= 0.0051
epoch:603, Train: Absolute_Depth_loss= 0.0136, Contrastive_Depth_loss= 0.0051
epoch:604, Train: Absolute_Depth_loss= 0.0133, Contrastive_Depth_loss= 0.0051
epoch:605, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:606, Train: Absolute_Depth_loss= 0.0139, Contrastive_Depth_loss= 0.0052
epoch:607, Train: Absolute_Depth_loss= 0.0141, Contrastive_Depth_loss= 0.0052
epoch:608, Train: Absolute_Depth_loss= 0.0143, Contrastive_Depth_loss= 0.0052
epoch:609, Train: Absolute_Depth_loss= 0.0148, Contrastive_Depth_loss= 0.0053
epoch:610, Train: Absolute_Depth_loss= 0.0137, Contrastive_Depth_loss= 0.0051
epoch:611, Train: Absolute_Depth_loss= 0.0134, Contrastive_Depth_loss= 0.0052
epoch:612, Train: Absolute_Depth_loss= 0.0143, Contrastive_Depth_loss= 0.0052
epoch:613, Train: Absolute_Depth_loss= 0.0144, Contrastive_Depth_loss= 0.0053
epoch:614, Train: Absolute_Depth_loss= 0.0139, Contrastive_Depth_loss= 0.0051
epoch:615, Train: Absolute_Depth_loss= 0.0147, Contrastive_Depth_loss= 0.0053
epoch:616, Train: Absolute_Depth_loss= 0.0141, Contrastive_Depth_loss= 0.0052
epoch:617, Train: Absolute_Depth_loss= 0.0142, Contrastive_Depth_loss= 0.0053
epoch:618, Train: Absolute_Depth_loss= 0.0127, Contrastive_Depth_loss= 0.0050
epoch:619, Train: Absolute_Depth_loss= 0.0131, Contrastive_Depth_loss= 0.0051
epoch:620, Train: Absolute_Depth_loss= 0.0131, Contrastive_Depth_loss= 0.0051
epoch:621, Train: Absolute_Depth_loss= 0.0137, Contrastive_Depth_loss= 0.0052
epoch:621, Val: val_threshold= 0.0006, val_ACC= 0.5271, val_ACER= 0.4839
epoch:621, Test: ACC= 0.5349, APCER= 0.5075, BPCER= 0.3158, ACER= 0.4116
epoch:622, Train: Absolute_Depth_loss= 0.0137, Contrastive_Depth_loss= 0.0052
epoch:623, Train: Absolute_Depth_loss= 0.0126, Contrastive_Depth_loss= 0.0050
epoch:624, Train: Absolute_Depth_loss= 0.0135, Contrastive_Depth_loss= 0.0051
epoch:625, Train: Absolute_Depth_loss= 0.0141, Contrastive_Depth_loss= 0.0052
epoch:626, Train: Absolute_Depth_loss= 0.0133, Contrastive_Depth_loss= 0.0051
epoch:627, Train: Absolute_Depth_loss= 0.0135, Contrastive_Depth_loss= 0.0052
epoch:628, Train: Absolute_Depth_loss= 0.0133, Contrastive_Depth_loss= 0.0051
epoch:629, Train: Absolute_Depth_loss= 0.0138, Contrastive_Depth_loss= 0.0052
epoch:630, Train: Absolute_Depth_loss= 0.0137, Contrastive_Depth_loss= 0.0052
epoch:631, Train: Absolute_Depth_loss= 0.0138, Contrastive_Depth_loss= 0.0052
epoch:632, Train: Absolute_Depth_loss= 0.0130, Contrastive_Depth_loss= 0.0051
epoch:633, Train: Absolute_Depth_loss= 0.0131, Contrastive_Depth_loss= 0.0051
epoch:634, Train: Absolute_Depth_loss= 0.0128, Contrastive_Depth_loss= 0.0051
epoch:635, Train: Absolute_Depth_loss= 0.0125, Contrastive_Depth_loss= 0.0050
epoch:636, Train: Absolute_Depth_loss= 0.0129, Contrastive_Depth_loss= 0.0051
epoch:637, Train: Absolute_Depth_loss= 0.0139, Contrastive_Depth_loss= 0.0051
epoch:638, Train: Absolute_Depth_loss= 0.0126, Contrastive_Depth_loss= 0.0050
epoch:639, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0050
epoch:640, Train: Absolute_Depth_loss= 0.0127, Contrastive_Depth_loss= 0.0050
epoch:641, Train: Absolute_Depth_loss= 0.0134, Contrastive_Depth_loss= 0.0051
epoch:641, Val: val_threshold= 0.0005, val_ACC= 0.5039, val_ACER= 0.4975
epoch:641, Test: ACC= 0.4070, APCER= 0.6479, BPCER= 0.3333, ACER= 0.4906
epoch:642, Train: Absolute_Depth_loss= 0.0130, Contrastive_Depth_loss= 0.0051
epoch:643, Train: Absolute_Depth_loss= 0.0131, Contrastive_Depth_loss= 0.0051
epoch:644, Train: Absolute_Depth_loss= 0.0124, Contrastive_Depth_loss= 0.0050
epoch:645, Train: Absolute_Depth_loss= 0.0122, Contrastive_Depth_loss= 0.0050
epoch:646, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0049
epoch:647, Train: Absolute_Depth_loss= 0.0125, Contrastive_Depth_loss= 0.0050
epoch:648, Train: Absolute_Depth_loss= 0.0123, Contrastive_Depth_loss= 0.0050
epoch:649, Train: Absolute_Depth_loss= 0.0126, Contrastive_Depth_loss= 0.0051
epoch:650, Train: Absolute_Depth_loss= 0.0126, Contrastive_Depth_loss= 0.0050
epoch:651, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0050
epoch:652, Train: Absolute_Depth_loss= 0.0121, Contrastive_Depth_loss= 0.0050
epoch:653, Train: Absolute_Depth_loss= 0.0120, Contrastive_Depth_loss= 0.0050
epoch:654, Train: Absolute_Depth_loss= 0.0120, Contrastive_Depth_loss= 0.0049
epoch:655, Train: Absolute_Depth_loss= 0.0122, Contrastive_Depth_loss= 0.0050
epoch:656, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0049
epoch:657, Train: Absolute_Depth_loss= 0.0128, Contrastive_Depth_loss= 0.0051
epoch:658, Train: Absolute_Depth_loss= 0.0121, Contrastive_Depth_loss= 0.0049
epoch:659, Train: Absolute_Depth_loss= 0.0120, Contrastive_Depth_loss= 0.0050
epoch:660, Train: Absolute_Depth_loss= 0.0123, Contrastive_Depth_loss= 0.0049
epoch:661, Train: Absolute_Depth_loss= 0.0124, Contrastive_Depth_loss= 0.0050
epoch:661, Val: val_threshold= 0.0008, val_ACC= 0.6279, val_ACER= 0.3868
epoch:661, Test: ACC= 0.4767, APCER= 0.5441, BPCER= 0.4444, ACER= 0.4943
epoch:662, Train: Absolute_Depth_loss= 0.0121, Contrastive_Depth_loss= 0.0049
epoch:663, Train: Absolute_Depth_loss= 0.0123, Contrastive_Depth_loss= 0.0050
epoch:664, Train: Absolute_Depth_loss= 0.0120, Contrastive_Depth_loss= 0.0049
epoch:665, Train: Absolute_Depth_loss= 0.0122, Contrastive_Depth_loss= 0.0049
epoch:666, Train: Absolute_Depth_loss= 0.0118, Contrastive_Depth_loss= 0.0049
epoch:667, Train: Absolute_Depth_loss= 0.0118, Contrastive_Depth_loss= 0.0050
epoch:668, Train: Absolute_Depth_loss= 0.0121, Contrastive_Depth_loss= 0.0050
epoch:669, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0049
epoch:670, Train: Absolute_Depth_loss= 0.0128, Contrastive_Depth_loss= 0.0051
epoch:671, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0048
epoch:672, Train: Absolute_Depth_loss= 0.0130, Contrastive_Depth_loss= 0.0051
epoch:673, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0049
epoch:674, Train: Absolute_Depth_loss= 0.0124, Contrastive_Depth_loss= 0.0050
epoch:675, Train: Absolute_Depth_loss= 0.0115, Contrastive_Depth_loss= 0.0049
epoch:676, Train: Absolute_Depth_loss= 0.0123, Contrastive_Depth_loss= 0.0049
epoch:677, Train: Absolute_Depth_loss= 0.0125, Contrastive_Depth_loss= 0.0050
epoch:678, Train: Absolute_Depth_loss= 0.0114, Contrastive_Depth_loss= 0.0049
epoch:679, Train: Absolute_Depth_loss= 0.0118, Contrastive_Depth_loss= 0.0049
epoch:680, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0049
epoch:681, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0048
epoch:681, Val: val_threshold= 0.0005, val_ACC= 0.6434, val_ACER= 0.3637
epoch:681, Test: ACC= 0.5233, APCER= 0.4783, BPCER= 0.4706, ACER= 0.4744
epoch:682, Train: Absolute_Depth_loss= 0.0117, Contrastive_Depth_loss= 0.0049
epoch:683, Train: Absolute_Depth_loss= 0.0123, Contrastive_Depth_loss= 0.0050
epoch:684, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0048
epoch:685, Train: Absolute_Depth_loss= 0.0121, Contrastive_Depth_loss= 0.0049
epoch:686, Train: Absolute_Depth_loss= 0.0118, Contrastive_Depth_loss= 0.0049
epoch:687, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0048
epoch:688, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0048
epoch:689, Train: Absolute_Depth_loss= 0.0115, Contrastive_Depth_loss= 0.0049
epoch:690, Train: Absolute_Depth_loss= 0.0117, Contrastive_Depth_loss= 0.0049
epoch:691, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0049
epoch:692, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0049
epoch:693, Train: Absolute_Depth_loss= 0.0115, Contrastive_Depth_loss= 0.0049
epoch:694, Train: Absolute_Depth_loss= 0.0108, Contrastive_Depth_loss= 0.0047
epoch:695, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0048
epoch:696, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0049
epoch:697, Train: Absolute_Depth_loss= 0.0115, Contrastive_Depth_loss= 0.0048
epoch:698, Train: Absolute_Depth_loss= 0.0119, Contrastive_Depth_loss= 0.0050
epoch:699, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0048
epoch:700, Train: Absolute_Depth_loss= 0.0103, Contrastive_Depth_loss= 0.0047
epoch:701, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0049
epoch:701, Val: val_threshold= 0.0005, val_ACC= 0.5194, val_ACER= 0.4871
epoch:701, Test: ACC= 0.4302, APCER= 0.5811, BPCER= 0.5000, ACER= 0.5405
epoch:702, Train: Absolute_Depth_loss= 0.0113, Contrastive_Depth_loss= 0.0048
epoch:703, Train: Absolute_Depth_loss= 0.0108, Contrastive_Depth_loss= 0.0048
epoch:704, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0048
epoch:705, Train: Absolute_Depth_loss= 0.0110, Contrastive_Depth_loss= 0.0047
epoch:706, Train: Absolute_Depth_loss= 0.0109, Contrastive_Depth_loss= 0.0048
epoch:707, Train: Absolute_Depth_loss= 0.0109, Contrastive_Depth_loss= 0.0047
epoch:708, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0048
epoch:709, Train: Absolute_Depth_loss= 0.0114, Contrastive_Depth_loss= 0.0049
epoch:710, Train: Absolute_Depth_loss= 0.0108, Contrastive_Depth_loss= 0.0047
epoch:711, Train: Absolute_Depth_loss= 0.0108, Contrastive_Depth_loss= 0.0047
epoch:712, Train: Absolute_Depth_loss= 0.0105, Contrastive_Depth_loss= 0.0047
epoch:713, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0047
epoch:714, Train: Absolute_Depth_loss= 0.0109, Contrastive_Depth_loss= 0.0047
epoch:715, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0049
epoch:716, Train: Absolute_Depth_loss= 0.0113, Contrastive_Depth_loss= 0.0048
epoch:717, Train: Absolute_Depth_loss= 0.0107, Contrastive_Depth_loss= 0.0048
epoch:718, Train: Absolute_Depth_loss= 0.0109, Contrastive_Depth_loss= 0.0047
epoch:719, Train: Absolute_Depth_loss= 0.0105, Contrastive_Depth_loss= 0.0047
epoch:720, Train: Absolute_Depth_loss= 0.0104, Contrastive_Depth_loss= 0.0047
epoch:721, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0048
epoch:721, Val: val_threshold= 0.0006, val_ACC= 0.5194, val_ACER= 0.4876
epoch:721, Test: ACC= 0.4535, APCER= 0.6027, BPCER= 0.2308, ACER= 0.4168
epoch:722, Train: Absolute_Depth_loss= 0.0110, Contrastive_Depth_loss= 0.0047
epoch:723, Train: Absolute_Depth_loss= 0.0109, Contrastive_Depth_loss= 0.0048
epoch:724, Train: Absolute_Depth_loss= 0.0097, Contrastive_Depth_loss= 0.0047
epoch:725, Train: Absolute_Depth_loss= 0.0106, Contrastive_Depth_loss= 0.0047
epoch:726, Train: Absolute_Depth_loss= 0.0112, Contrastive_Depth_loss= 0.0048
epoch:727, Train: Absolute_Depth_loss= 0.0106, Contrastive_Depth_loss= 0.0046
epoch:728, Train: Absolute_Depth_loss= 0.0104, Contrastive_Depth_loss= 0.0047
epoch:729, Train: Absolute_Depth_loss= 0.0101, Contrastive_Depth_loss= 0.0046
epoch:730, Train: Absolute_Depth_loss= 0.0103, Contrastive_Depth_loss= 0.0046
epoch:731, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0047
epoch:732, Train: Absolute_Depth_loss= 0.0102, Contrastive_Depth_loss= 0.0047
epoch:733, Train: Absolute_Depth_loss= 0.0111, Contrastive_Depth_loss= 0.0048
epoch:734, Train: Absolute_Depth_loss= 0.0106, Contrastive_Depth_loss= 0.0047
epoch:735, Train: Absolute_Depth_loss= 0.0106, Contrastive_Depth_loss= 0.0047
epoch:736, Train: Absolute_Depth_loss= 0.0116, Contrastive_Depth_loss= 0.0048
epoch:737, Train: Absolute_Depth_loss= 0.0114, Contrastive_Depth_loss= 0.0048
epoch:738, Train: Absolute
|
|
It seems that many people have problems with training. As we all know, data augmentation is very vital to train models.
Then, you may reproduce the paper results. |
|
Thank you so much for your reply sir,
I have removed the frame selection part from the code because I have taken
only one frame of a video and generated the depth map and then given it to
the CDCn model
All other preprocessing is the same as the source code sir
So if possible can you able to share your code? if possible
it will more useful for me to compare
because I have only very little time to graduate, I found the mistake just
on this day. really sorry to trouble you sir
Hope for a reply sir (life player)
…On Mon, 31 May 2021 at 17:16, lifeplayer ***@***.***> wrote:
It seems that many people have problems with training. As we all know,
data augmentation is very vital to train models.
For this code, we may ignore some detail when we modify the code.
- In terms of data loader design, the author randomly selects a frame
from each video in the training set for each epoch， and it may slow down
training progress ;
- It uses multi-scale data augmentation in the operation of cropping
the face area.
Then, you may reproduce the paper results.
This link is one of my pre-trained models, you can try to finetune to get
the paper result.
https://drive.google.com/file/d/1RmxiWzXvXgMV6VQG3oA8u1XafftQW5QC/view?usp=sharing
————————————————————————————————————————
About the Oulu dataset, I'm sorry I can‘t provide it now.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B27JV7MLQM4PNLEE2E3TQNHYJANCNFSM4LVPHJTA>
.
|
|
Only one frame extracted from one video？Maybe training data too little. It’s not convenient for me to share code recently. Or you can provide your social network account for us to communicate privately. |
thanks for reply! I can wait until it's convenient for you, and hope I've got it then, |
|
my wechat : spv_krv can you please share yours, so we can discuss there |
I have emailed you. PLZ CHECK again. |
I'm sorry to bother you again! but if it's convenient for you,can you share me the dataset?I haven't got the dataset yet! 0.o |
Hi bro, Thanks in advance |
@xiao-keeplearning can you help me answer above questions? |
Did you train the model successfully? Can you share me some experiences? Thanks in advance |
|
Hi,
Here the Testing is carried out with the help of val_threshold,
sorry, I don't have Replace attack dataset, so don't have any idea about it
For Me also very difficult to obtain the paper result
my best wishes,
…On Thu, 10 Jun 2021 at 14:14, luan1412167 ***@***.***> wrote:
@xiao-keeplearning <https://github.com/xiao-keeplearning>
my wechat: spv_krv
can you please share yours, so we can discuss there
Did you train the model successfully? Can you share me some experiences?
I'm training the network on replay-attack dataset. Something got me
confused.
1, What is val_threshold? it is so small in validation, as me know, it
equal to 0 (spoof) and 1 (live).
2. val_ACC, val_ACER is not stable and not good. Did you reproduce a good
result with source code in this repo?
Thanks in advance
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B2YOIRLGXIUNGD5DMD3TSBJ6DANCNFSM4LVPHJTA>
.
|
thanks for your response |
|
Yup, ...
I will converge after 800 epoch
…On Fri, Jun 11, 2021, 1:29 PM luan1412167 ***@***.***> wrote:
Hi, Here the Testing is carried out with the help of val_threshold, sorry,
I don't have Replace attack dataset, so don't have any idea about it For Me
also very difficult to obtain the paper result my best wishes,
… <#m_1236839607717487786_>
On Thu, 10 Jun 2021 at 14:14, luan1412167 *@*.***> wrote:
@xiao-keeplearning <https://github.com/xiao-keeplearning>
https://github.com/xiao-keeplearning my wechat: spv_krv can you please
share yours, so we can discuss there Did you train the model successfully?
Can you share me some experiences? I'm training the network on
replay-attack dataset. Something got me confused. 1, What is val_threshold?
it is so small in validation, as me know, it equal to 0 (spoof) and 1
(live). 2. val_ACC, val_ACER is not stable and not good. Did you reproduce
a good result with source code in this repo? Thanks in advance — You are
receiving this because you were mentioned. Reply to this email directly,
view it on GitHub <#6 (comment)
<#6 (comment)>>, or
unsubscribe
https://github.com/notifications/unsubscribe-auth/AMM7B2YOIRLGXIUNGD5DMD3TSBJ6DANCNFSM4LVPHJTA
.
thanks for your response
I find that the author tell CDCNpp converge slow, so Did you have train
about 600-900 epochs? Is it converge?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AMM7B277IY5YOF6LV6FH7PLTSGNLBANCNFSM4LVPHJTA>
.
|
|
Oh, |
@xiao-keeplearning @Coderx7 @ngoanpv @marvelyou @22wei22 @Prasandhmcw @Coderx7 @xiao-keeplearning @laoxing797373 @MNYxin @ZitongYu @lz28lz28 Auc_Valu 0.5181873376646298 Still unable to reach the paper result, I have used the pre-train model you have shared Can you please help with this It will be a great help Thanks in advance |
i have attached the test map score as follows 0.0028912031557410955 0 |
|
@punitha-valli @xiao-keeplearning Could you share me the OULU dataset. I need this dataset for completing an idea in my paper of cross domain face anti spoofing. My email is huyhung.dknec@gmail.com. Thank you so much. |
Excuse me, I have a question about the clipping operation. Can a full depth map still serve as monitoring information if the image is cropped randomly? |
|
@xiao-keeplearning Sorry to bother you. Could you tell me how many frames did you use from 1 video? I found that using PRnet to generate depth_map for every frame really takes too much time... |
|
@jamesdongdong My experience is to sample 3-6 frames in 1 second |
@xiao-keeplearning Thanks for your advice! BTW, did you remember how long PRnet took to generate 1 depth map? |
|
@jamesdongdong It's too slow. I used the 3ddfa project code to generate roughly 10 depth maps in 1s on a 1080Ti. |
|
@xiao-keeplearning |
IMO 0 is real and 1 is spoof? |
|
@luan1412167 can you share OULU-NPU dataset for me through phamkhactu98@gmail.com. I have on end of master thesis. |
and others
When I train on oulu Proro1, I find the loss hard to converge.
Here is part of my train log.The ACER fluctuates near 5%, though I have tried to adjust lr.
Could you share your train.log or provide any advice?
The text was updated successfully, but these errors were encountered: