karfly/learnable-triangulation-pytorch

Issues

• Pretrained weights file for CMU Dataset

  #72 opened 5 years ago by Samleo8
  2

• Pretaining for 2D Backbone, with CMU Dataset

  #77 opened 5 years ago by Samleo8
  1

• CMU-trained Weights and Cross dataset evaluation

  #83 opened 4 years ago by Samleo8
  5

• Is it possible to use this excellent tool without any training?

  #52 opened 5 years ago by annihi1ation
  3

• Can it run in realtime?

  #97 opened 4 years ago by dandevs
  3

• Use of an own data set, taken from 2 equal webcams with different positions

  #67 opened 4 years ago by Gowan1998
  11

• How does the number of camera views influence the performance on Human 3.6m dataset?

  #95 opened 4 years ago by huagl
  6

• Evaluating without Human3.6M dataset?

  #71 opened 4 years ago by BartArtanim
  4

• Bug with n_joints in `train.py`

  #73 opened 4 years ago by Samleo8
  3

• Using model on single-view videos (or webcam input)

  #78 opened 4 years ago by SynStratos
  2

• preprocess the dataset:Cannot load CDF C library; checked . Try 'os.environ["CDF_LIB"] = library_directory'

  #84 opened 4 years ago by angechen
  1

• GPU utility during training

  #103 opened 4 years ago by kyang-06
  2

• Viewing data in 3D space

  #102 opened 4 years ago by jamalknight
  1

• Calculating MPJPE

  #98 opened 4 years ago by jamalknight
  1

• H36M Test set

  #101 opened 4 years ago by jamalknight
  2

• Camera information

  #99 opened 4 years ago by jamalknight
  2

• Testing

  #100 opened 4 years ago by jamalknight
  1

• Issue running evaluation

  #96 opened 4 years ago by jamalknight
  2

• How to stop training?

  #94 opened 4 years ago by jamalknight
  2

• Question: Training duration (volumetric)

  #90 opened 4 years ago by jamalknight
  1

• How many iters for training?

  #92 opened 4 years ago by jamalknight
  6

• When fill retval['cameras'], why do transpose with T in camera params?

  #93 opened 4 years ago by dulibubai
  2

• Does the number of views have to be the same for training and evaluation?

  #91 opened 4 years ago by cherryjm
  0

• Issues, notes and documentation while testing on the CMU dataset, using volumetric model

  #75 opened 5 years ago by Samleo8
  14

• When training, CUDA runs out of memory - How can I reduce the batch size?

  #89 opened 4 years ago by jamalknight
  2

• Question: Will pre processing all frames (instead of skipping frames) make it temporally smoother? ie. less jittering

  #88 opened 4 years ago by jamalknight
  2

• Error wrapping 3D keypoints, bounding boxes and camera intrinsics together

  #87 opened 4 years ago by jamalknight
  2

• Error running collect-bboxes.py

  #86 opened 4 years ago by jamalknight
  2

• Dropbox path not working

  #82 opened 4 years ago by jamalknight
  4

• Model can train from checkpoint but cannot continue training successively

  #81 opened 4 years ago by Samleo8
  2

• training time

  #69 opened 4 years ago by angechen
  4

• CUDA out of memory error (only after first epoch)

  #80 opened 4 years ago by Samleo8
  1

• Training of CMU dataset gets stuck on batch 1

  #79 opened 4 years ago by Samleo8
  5

• Problem with partial occluded bodies when evaluating on CMU dataset

  #76 opened 5 years ago by Samleo8
  0

• Failed to download the h36m.zip while doing the preprocessing (the 5th step)

  #68 opened 5 years ago by DiffenYu
  2

• How to interpret human36m-multiview-labels-GTbboxes.npy?

  #70 opened 5 years ago by wwang107
  6

• Some problems about the dataset Human36M

  #63 opened 5 years ago by sunny0212452
  6

• Possible bug in update_after_resize

  #65 opened 5 years ago by ThomasOerkild
  1

• About training and evaluating in a single view setting when using Volumetric Triangulation

  #64 opened 5 years ago by ShirleyMaxx
  4

• Use the test data of dataset Human36M

  #59 opened 5 years ago by sunny0212452
  1

• Regarding the volumetric triangulation, how is the heatmap projected into the Volume

  #57 opened 5 years ago by davidmaeurer
  5

• CMU hand keypoints dataset using learnable triangulation

  #61 opened 5 years ago by pablovela5620
  1

• Questions about training overfit on human6m dataset

  #62 opened 5 years ago by sunny0212452
  1

• How to train weights in confidence branch?

  #60 opened 5 years ago by jevonswang
  1

• Are you going to use MoVi dataset?

  #58 opened 5 years ago by Ongoza
  1

• Visualising the evaluation results

  #51 opened 5 years ago by Samleo8
  10

• backbone pretraining details

  #56 opened 5 years ago by pablovela5620
  4

• details of experimental settings on CMU dataset

  #55 opened 5 years ago by MasakazuYposhimura
  2

• CMU Panoptic style annotation

  #53 opened 5 years ago by isarandi
  4

• Bbox requirements - Do they need to be square?

  #54 opened 5 years ago by Samleo8
  1