yinghdb/EmbedMask

Issues

• Why margin need to be scaled by "scale_size"?

  #15 opened 5 years ago by shoutOutYangJie
  9

• New version

  #18 opened 4 years ago by JyunYuLai
  2

• Positive samples for proposal embedding and margin

  #17 opened 4 years ago by JyunYuLai
  2

• focal loss parameters

  #16 opened 5 years ago by gewenbin292
  1

• Question about lovasz_hinge_loss

  #14 opened 5 years ago by Epiphqny
  2

• Embedding Visualization

  #13 opened 5 years ago by eugenelawrence
  1

• Speed performance

  #12 opened 5 years ago by Epiphqny
  2

• reformulate embedmask based on detectron2

  #9 opened 5 years ago by TengFeiHan0
  11

• About comparison with state-of-the-art methods

  #11 opened 5 years ago by mychina75
  3

• Bugs in get_pos_proposal_indexes for mask loss

  #10 opened 5 years ago by Devin147
  1

• about smooth_loss in training

  #8 opened 5 years ago by UdonDa
  2

• unreachable gradient

  #7 opened 5 years ago by afantideng
  5

• Performance

  #5 opened 5 years ago by trungpham2606
  7

• Smooth loss is 0.0000 when Training with one GPU

  #6 opened 5 years ago by SkeletonOne
  2

• Nan for classification loss in training stage.

  #4 opened 5 years ago by wangyuxin87
  1

• FCOS structure

  #3 opened 5 years ago by trungpham2606
  3

• Train Speed

  #1 opened 5 years ago by RainHxj
  1

• Why so many outputs

  #2 opened 5 years ago by lucasjinreal
  0