kwea123/CasMVSNet_pl

Issues

• 自采数据深度图估计效果很差

  #21 opened 4 years ago by aweijx
  13

• Reference image

  #52 opened 9 months ago by mumuliu66
  0

• Point cloud is worse than original project on DTU

  #37 opened 2 years ago by YuhsiHu
  23

• Creating my own datasets

  #50 opened 2 years ago by Irina51244
  0

• the loss and train_abs_err=nan

  #49 opened 2 years ago by hakeemray
  0

• 您好，请问配置docker时出现以下问题是因为什么呢？

  #48 opened 2 years ago by FrozenCo
  0

• multi gpu mode is much slower than single gpu mode

  #31 opened 4 years ago by tau-yihouxiang
  3

• What is the result on testing sets like TNT and ETH3D when the model is trained on BlendedMVS

  #47 opened 3 years ago by NoOneUST
  0

• Test on Tanks and Temples

  #33 opened 4 years ago by caoPhoenix
  5

• Why are depth intervals hand tuned but not read from the camera files?

  #46 opened 3 years ago by vicissitude1999
  0

• debugging with pdb

  #44 opened 3 years ago by wtyuan96
  1

• about training problem

  #45 opened 3 years ago by alsilence
  0

• Blended checkpoint (15) worse performance than DTU checkpoint (10) on smaller objects

  #42 opened 3 years ago by eddienewton
  0

• 能提供一下Advanced数据集的相机参数吗？

  #41 opened 3 years ago by zhaozj19
  1

• Results on DTU

  #40 opened 3 years ago by stillwang96
  1

• question about purpose of scale_factor in blendedmvs.py

  #39 opened 3 years ago by LuckyOne09
  2

• Training error on BlendedMVS dataset.

  #38 opened 3 years ago by saedrna
  3

• Training loss

  #32 opened 3 years ago by YuhsiHu
  1

• ImportError when run eval.py

  #36 opened 3 years ago by YuhsiHu
  1

• Is there any pretrained model?

  #35 opened 3 years ago by double-yang-96
  1

• Calculate real depth

  #34 opened 4 years ago by katealenic
  1

• Evaluate on DTU

  #30 opened 4 years ago by NoOneUST
  1

• A question when test on dtu

  #29 opened 4 years ago by LuckyOne09
  2

• small mistakes in /datasets/tanks.py I think

  #28 opened 4 years ago by LuckyOne09
  5

• Testing on DTU

  #25 opened 4 years ago by NoOneUST
  1

• About Homography Warping

  #23 opened 4 years ago by GONGYue1231
  2

• How to remove the InplaceABN part

  #27 opened 4 years ago by NoOneUST
  1

• How should we test the model on ETH3D dataset？

  #26 opened 4 years ago by NoOneUST
  1

• Tanks results of casmvsnet_pl is worse than the original casmvsnet

  #14 opened 4 years ago by chaytonmin
  8

• About the CasMVSNet

  #2 opened 4 years ago by whubaichuan
  9

• Testing with colmap poses

  #13 opened 4 years ago by cvgogogo
  2

• About Inplace-abn

  #24 opened 4 years ago by Tangshengku
  1

• How to run the code on Pytorch version>=1.6?

  #22 opened 4 years ago by NoOneUST
  1

• About the Inplace-ABN

  #20 opened 4 years ago by XYZ-qiyh
  1

• About the time of forward inference

  #19 opened 4 years ago by XYZ-qiyh
  1

• Choosing depth interval for Tanks and Temples dataset

  #18 opened 4 years ago by phongnhhn92
  2

• Does this method work on real-captured images ?

  #17 opened 4 years ago by phongnhhn92
  4

• Confidence map

  #16 opened 4 years ago by phongnhhn92
  1

• Fusing depth maps estimated for Tanks and Temples dataset on 2080Ti

  #15 opened 4 years ago by Txusheng
  2

• testing on tanks datasets

  #12 opened 5 years ago by chaytonmin
  2

• About homography warping

  #11 opened 5 years ago by chaytonmin
  4

• Increasing point generation for views with varying depths

  #10 opened 5 years ago by KevinCain
  2

• Size mismatch when running with release model trained on BlendedMVS data

  #8 opened 5 years ago by KevinCain
  2

• Zero points in output .ply

  #9 opened 5 years ago by KevinCain
  6

• Inference from DTU data

  #7 opened 5 years ago by KevinCain
  6

• evaluation metric

  #6 opened 5 years ago by cvgogogo
  4

• the influence of batchsize with Inplace-ABN

  #5 opened 5 years ago by chaytonmin
  5

• the error of the current depth fusion codes

  #4 opened 5 years ago by chaytonmin
  2

• Pretrained BlendedMVS model

  #3 opened 5 years ago by KevinCain
  2