gudovskiy/cflow-ad

Issues

• About the difficulties of exporting onnx

  #32 opened 2 years ago by nullhd2
  10

• test loss

  #46 opened 2 years ago by MARS-J
  1

• How to check visulization ?

  #50 opened a year ago by Abhijeet241093
  0

• NameError: name 'mobilenet_v3_large' is not defined

  #49 opened a year ago by Abhijeet241093
  0

• Custom datasets

  #27 opened 3 years ago by mouxinyue1
  9

• Why use positional embedding as conditional vector?

  #43 opened 2 years ago by eric-tc
  2

• The motivation and mechanism of conditional flow

  #48 opened a year ago by bearinsuke
  0

• value of N

  #44 opened 2 years ago by Howie86
  0

• Why using meta-epoch training paradigm

  #18 opened 3 years ago by DonaldRR
  8

• Increase hyperparameter N

  #19 opened 3 years ago by rvermeire
  5

• Huge Training Loss after a certain number of epochs

  #36 opened 2 years ago by aritra3520
  1

• Error when use resnet50 as feature extractor

  #30 opened 2 years ago by Howeng98
  3

• Did the encoder training parameters setting in the paper refer to the pretrain parameters setting of the encoder?

  #42 opened 2 years ago by Howie86
  0

• Did the encoder training parameters setting in the paper refer to the pretrain parameters setting of the encoder?

  #41 opened 2 years ago by Howie86
  2

• What is the B value here in extracting feature map

  #40 opened 2 years ago by ThinhNguyenHoang
  2

• Same model to inspect different but similar objects.

  #35 opened 2 years ago by Tekno-H
  0

• U-L2?

  #34 opened 2 years ago by mjack3
  1

• Tune parameters to get best results.

  #31 opened 2 years ago by Tekno-H
  2

• Calculate detection AUROC from anomaly map

  #29 opened 2 years ago by Howeng98
  1

• Inference speed increase with the time

  #28 opened 3 years ago by mjack3
  1

• heat map visualization

  #7 opened 3 years ago by OzanErcan6
  5

• heatmap

  #26 opened 3 years ago by Dyy-1214
  0

• Improper normalization of the scores?

  #24 opened 3 years ago by marco-rudolph
  9

• Detection AUROC and segmentation AUROC behavior

  #25 opened 3 years ago by mjack3
  1

• Negative Log Sigmoid Log Likelihood Loss?

  #23 opened 3 years ago by marco-rudolph
  2

• Calculate detection AUROC

  #22 opened 3 years ago by jpmrs1313
  1

• RuntimeError: Cannot insert a Tensor that requires grad as a constant. Consider making it a parameter or input, or detaching the gradient

  #21 opened 3 years ago by volkov-maxim
  2

• About detection speed

  #20 opened 3 years ago by xfby2016
  3

• Why num channels is divided here?

  #17 opened 3 years ago by mysephi
  1

• How many GPU hours are required?

  #16 opened 3 years ago by caoyunkang
  3

• Consuming a lot of memory when calculating segmentation AUROC

  #15 opened 3 years ago by gaowq2017
  1

• The speed of train so slowly

  #14 opened 3 years ago by gaowq2017
  0

• test score_label Inconsistent when test images Inconsistent in Same image

  #13 opened 3 years ago by montensorrt
  2

• Why did you set $condition_vec=128$ not the simplest $condition_vec=4$ ?

  #12 opened 3 years ago by hihunjin
  2

• how to inference single image?

  #11 opened 3 years ago by gewenpulan
  6

• Is it possible to get a class name of a surface/object on inference?

  #10 opened 3 years ago by volkov-maxim
  1

• Is it possible to get a class name of an anomaly on inference?

  #9 opened 3 years ago by volkov-maxim
  1

• Under what license the project is?

  #8 opened 3 years ago by volkov-maxim
  1

• IoU Evaluation

  #6 opened 3 years ago by ZeWang95
  1

• Large test loss, and thresholds

  #4 opened 3 years ago by letmejoin
  5

• WRN50 Transistor AUC97.99 ?

  #5 opened 3 years ago by coordxyz
  1

• Question about image resizing and evaluation methods in localization task

  #3 opened 3 years ago by shot1107
  2

• why taking dataloader and training times very slow?

  #2 opened 3 years ago by ingbeeedd
  2

• where is checkpoint? pretrained weights to reproduce your results?

  #1 opened 3 years ago by wonchul-kim
  4