princeton-vl/DeepV2D

Issues

• Blas xGEMMBatched launch failed : a.shape=[7,3,3], b.shape=[7,3,1], m=3, n=1, k=3, batch_size=7

  #21 opened 4 years ago by duanzhimin14
  2

• Dynamic Frames For Inference

  #33 opened a year ago by jytime
  1

• Any idea to release a pytorch version？

  #40 opened 4 years ago by sconlyshootery
  3

• LS-OPTIMIZATION LAYER back propagation

  #49 opened 2 years ago by waveajay
  0

• Can't run demo with batch size 8

  #10 opened 5 years ago by roxanneluo
  5

• Why the memory keeps increasing during training

  #48 opened 3 years ago by myalos
  0

• kitti.ckpt does not support global mode

  #39 opened 4 years ago by weihaosky
  1

• Why here self.total_loss is set to depth_loss

  #47 opened 4 years ago by sconlyshootery
  0

• demo_slam on custom video?

  #35 opened 4 years ago by sathishkumark27
  1

• How to inference on a longer video?

  #46 opened 4 years ago by KDUserStudy
  0

• About ScanNet

  #44 opened 4 years ago by LifeBeyondExpectations
  6

• about jacobian

  #45 opened 4 years ago by sconlyshootery
  0

• Results even better using "default" validation method?

  #6 opened 5 years ago by cgebbe
  2

• Why img is flip before resize

  #43 opened 4 years ago by sconlyshootery
  2

• Image reconstruction quality on kitti

  #42 opened 4 years ago by TWJianNuo
  0

• Why Depth prediciton on kitti dataset is a relative depth

  #41 opened 4 years ago by TWJianNuo
  0

• cannot load checkpoint

  #13 opened 5 years ago by l0stpenguin
  2

• Question about pose preprocessing

  #38 opened 4 years ago by jwh97nn
  1

• cuda version error?

  #26 opened 4 years ago by apxlwl
  0

• Why are the intrinsic params divided by 4?

  #37 opened 4 years ago by ootts
  0

• How did you evaluate TUM using translational rmse(m/s)

  #9 opened 5 years ago by eric-yyjau
  3

• Dose demo_slam on nyu support loop closure?

  #36 opened 4 years ago by sathishkumark27
  0

• How to implement a single-view demo?

  #34 opened 4 years ago by Mofafa
  1

• ScanNet training

  #32 opened 4 years ago by ChenHoy
  5

• Questions about intrinsics for ScanNet

  #30 opened 4 years ago by MathLens
  2

• Question about scaling?

  #31 opened 4 years ago by ryancll
  1

• How to load 2-stage checkpoints for evaluation?

  #29 opened 4 years ago by ryancll
  1

• Output of Demo scripts

  #25 opened 4 years ago by aakash26
  3

• About heavy distortion of video.

  #28 opened 4 years ago by shuluoshu
  1

• ImportError: No module named 'vtkOpenGLKitPython'

  #27 opened 4 years ago by ghoshaw
  0

• NYU preprocessing

  #24 opened 4 years ago by apxlwl
  1

• NYUD pose

  #20 opened 4 years ago by apxlwl
  1

• code ambiguity

  #23 opened 4 years ago by mhiraoka
  2

• why scale depth_pred for evaluation?

  #22 opened 4 years ago by flamehaze1115
  3

• connecting error when downloading the pretrained model

  #19 opened 4 years ago by theodore-zxd
  2

• NYU gt association and pose

  #18 opened 4 years ago by YamamotoSayaka
  2

• Significance of multiplying the translation vector by 0.1(args['scale'])

  #15 opened 5 years ago by rabin9948
  7

• Is your method end-2-end ?

  #17 opened 5 years ago by phongnhhn92
  4

• scale in evaluation?

  #16 opened 5 years ago by rxqy
  2

• InvalidArgumentError (see above for traceback): Cholesky decomposition was not successful. The input might not be valid.

  #14 opened 5 years ago by hzxie
  0

• NYU Depth TFRecord

  #12 opened 5 years ago by charleshamesse
  0

• Test on video

  #11 opened 5 years ago by roxanneluo
  1

• tensorflow-gpu version

  #8 opened 5 years ago by eric-yyjau
  1

• Intrinsics file - what do numbers mean?

  #5 opened 5 years ago by cgebbe
  2

• question about view pooling

  #7 opened 5 years ago by MASTERCHIEF343
  0

• Simultaneous Tracking and Mapping?

  #4 opened 6 years ago by Yuliang-Zou
  7

• Problem running KITTI demo

  #3 opened 6 years ago by cgebbe
  2

• Have you tried Tracking and Mapping like DeepTAM?

  #2 opened 6 years ago by xy-guo
  0

• Problem running demo

  #1 opened 6 years ago by UannaFF
  0