neptune-ai/open-solution-salt-identification

Issues

• implement center_encoder l2 loss

  #49 opened 6 years ago by jakubczakon
  0

• experiment with dropout2d in the bottom layer

  #76 opened 6 years ago by jakubczakon
  0

• Implement PSPNet

  #87 opened 6 years ago by jakubczakon
  0

• Installation fails due to another conflicting pandas version

  #96 opened 5 years ago by NeolithEra
  3

• Conflicts between open-solution-salt-identification and ipython

  #95 opened 5 years ago by NeolithEra
  5

• OSError: [Errno 101] Network is unreachable

  #93 opened 6 years ago by OsloAI
  1

• Prepare data based on jigsaw puzzle

  #90 opened 6 years ago by jakubczakon
  0

• implement OCNet

  #86 opened 6 years ago by jakubczakon
  0

• experiment with loss schedules

  #85 opened 6 years ago by jakubczakon
  0

• second level model for optimal threshold estimation

  #77 opened 6 years ago by jakubczakon
  0

• Experiment with mean teacher

  #71 opened 6 years ago by jakubczakon
  0

• Implement feature pyramid with global attention

  #55 opened 6 years ago by jakubczakon
  0

• Implement SWA optimization

  #48 opened 6 years ago by jakubczakon
  0

• Implement post-processing block JOSE

  #46 opened 6 years ago by jakubczakon
  0

• Lovasz loss

  #94 opened 6 years ago by yuanqing811
  1

• index 0 is out of bounds for axis 0 with size 0

  #92 opened 6 years ago by OsloAI
  1

• IndexError: too many indices for array

  #91 opened 6 years ago by OsloAI
  1

• explore other pretrained

  #62 opened 6 years ago by jakubczakon
  0

• add scse blocks after encoders

  #88 opened 6 years ago by jakubczakon
  0

• implement replication y + reflection x

  #64 opened 6 years ago by jakubczakon
  0

• try attention + squeeze and excite

  #70 opened 6 years ago by jakubczakon
  0

• experiment with large kernel matters

  #83 opened 6 years ago by jakubczakon
  0

• explore wavelet representation

  #84 opened 6 years ago by jakubczakon
  0

• Implement constrained size loss

  #68 opened 6 years ago by jakubczakon
  0

• Explore gabor filters representation

  #43 opened 6 years ago by jakubczakon
  0

• Implement Large Kernels Matter

  #69 opened 6 years ago by jakubczakon
  0

• Memory Allocation of Solution 4

  #82 opened 6 years ago by xhh232018
  1

• Experiment with larger size

  #75 opened 6 years ago by jakubczakon
  0

• No module named 'neptune'

  #66 opened 6 years ago by qinhui99
  2

• Add synthetically generated small border masks

  #74 opened 6 years ago by jakubczakon
  0

• Implement adaptive-threshold evaluation checkpoint

  #72 opened 6 years ago by jakubczakon
  0

• Experiment with dilated targets for small objects

  #67 opened 6 years ago by jakubczakon
  0

• train with encoder LR smaller

  #53 opened 6 years ago by jakubczakon
  0

• CV for the best threshold

  #57 opened 6 years ago by jakubczakon
  0

• try mixed penaulty loss

  #56 opened 6 years ago by jakubczakon
  0

• explore border penaulty loss

  #58 opened 6 years ago by jakubczakon
  0

• explore focal loss

  #63 opened 6 years ago by jakubczakon
  0

• Increase unet resolution

  #51 opened 6 years ago by jakubczakon
  0

• explore squeeze and excitation block

  #60 opened 6 years ago by jakubczakon
  0

• Implement simple model averaging

  #61 opened 6 years ago by jakubczakon
  0

• explore replication padding in convolutions

  #59 opened 6 years ago by jakubczakon
  0

• try hypercolumn approach

  #54 opened 6 years ago by jakubczakon
  0

• Implement attention into decoders

  #47 opened 6 years ago by jakubczakon
  0

• TypeError: can't multiply sequence by non-int of type 'float' while training

  #52 opened 6 years ago by Diyago
  2

• Implement easy fine-tuning handling

  #45 opened 6 years ago by jakubczakon
  0

• Implement the lovasz softmax loss

  #42 opened 6 years ago by jakubczakon
  0

• implement batch norm in the decoder part

  #44 opened 6 years ago by jakubczakon
  0

• train on 128 + 32x2 pad

  #39 opened 6 years ago by jakubczakon
  0

• train with sgd

  #41 opened 6 years ago by jakubczakon
  0

• Experiment with random gradient

  #38 opened 6 years ago by jakubczakon
  0