NVlabs/ssn_superpixels

Issues

• extract the features of the superpixel area

  #38 opened a year ago by xwcUP
  0

• custom training

  #37 opened 4 years ago by Zampal94
  1

• Pytorch Implementation License

  #36 opened 4 years ago by andrewsonga
  0

• Superpixel border issue on BSDS500 dataset

  #35 opened 4 years ago by yuanqqq
  5

• about the label and loss

  #34 opened 4 years ago by nmvbxcz
  2

• Question of the function of spix_init

  #33 opened 4 years ago by yueyu-stu
  2

• I cannot understand “n*9” mentioned in the paper

  #29 opened 5 years ago by TakeruIto
  1

• Why is random cropping / patching and random scaling applied continuously throughout training?

  #31 opened 5 years ago
  1

• how to run the code???

  #32 opened 5 years ago by cyy1426
  1

• F0630 15:37:53.939426 12256 math_functions.cu:79] Check failed: error == cudaSuccess (74 vs. 0) misaligned address *** Check failure stack trace: ***

  #30 opened 5 years ago by ZubairKhan001
  0

• The implementation of pytorch version

  #28 opened 5 years ago by CYang0515
  6

• what does the training loss curve look like

  #27 opened 5 years ago
  5

• Some questions on source code

  #26 opened 5 years ago by Bobholamovic
  7

• The BSDS result from the pre-trained model does not match the paper one

  #11 opened 6 years ago
  13

• about Evaluation

  #25 opened 5 years ago by lokiii6
  1

• Some question about initialization of group conv layer 'concat_spixel_feat_50'

  #24 opened 5 years ago by yuzhegao
  3

• About some custom caffe layer ?

  #22 opened 5 years ago by yuzhegao
  3

• input image size different for training and testing on the BSDS500 dataset??

  #23 opened 5 years ago
  1

• where to install Caffe inside "ssn_superpixels"

  #21 opened 5 years ago
  2

• Some confusion about training with label

  #17 opened 5 years ago by forrestsz
  1

• some question of paper about training

  #20 opened 5 years ago by yuzhegao
  4

• Aboutn

  #19 opened 5 years ago by yuzhegao
  0

• Questions about semantic segmentation testing

  #18 opened 5 years ago by Yajha
  2

• Presence of hard association in the compactness portion of loss function

  #16 opened 5 years ago
  1

• caffe build configuration？

  #15 opened 5 years ago by WenqiangX
  1

• converge very slowly

  #13 opened 6 years ago by jiayusun
  2

• superpixel results

  #14 opened 6 years ago by zicair
  5

• Training with self-made dataset

  #12 opened 6 years ago by ShaneZHOU1994
  4

• About the weights of the loss function

  #9 opened 6 years ago
  4

• dataset

  #8 opened 6 years ago by Eileen2014
  1

• Is there any pytorch version of this work ?

  #10 opened 6 years ago by ahmedmazariML
  1

• Test for city_scapes dataset?

  #7 opened 6 years ago by Wanger-SJTU
  2

• h_grid, w_grid is not used

  #6 opened 6 years ago by iliadsouti
  1

• How can l get the features vector associated to each superpixel ?

  #2 opened 6 years ago by pinkfloyd06
  3

• optical flow with superpixel

  #1 opened 6 years ago by xiaofanglegoc
  4

• when i run your code using pycharm,the error is..

  #5 opened 6 years ago by farm365
  2

• Question About your paper

  #3 opened 6 years ago by lxtGH
  3

• optical flow network

  #4 opened 6 years ago by texastone
  2