GuyTevet/MotionCLIP

Issues

• different joint index for thetas(pose) and joints3d?

  #41 opened 2 months ago by meongeun
  0

• amass_30fps_db.pt file

  #40 opened 2 months ago by meongeun
  0

• Was (0, 0, 0) being passed as translation to the encoder during training?

  #39 opened 4 months ago by Ivan-Klabucar
  0

• What device did you use to train the model and how long did it cost?

  #38 opened 7 months ago by Croolch
  0

• Motions are static

  #15 opened 2 years ago by shunlinlu
  4

• ffmpeg version

  #37 opened a year ago by Tong-Hao741
  4

• AttributeError: 'AMASS' object has no attribute 'nfeats'

  #36 opened a year ago by DTOSCANO461
  2

• consistency of the motion encoder and the motion decoder

  #35 opened a year ago by dsdf783
  1

• Create conda environment failed

  #18 opened a year ago by ryanhoo
  1

• Visualize failed

  #20 opened a year ago by ShungJhon
  2

• 'amass_30fps_legacy_db.pt'

  #34 opened a year ago by dsdf783
  2

• AssertionError：data/amass/amass_30fps_legacy_db.pt

  #13 opened 2 years ago by ywdong
  2

• Can not find "data/amass/amass_30fps_legacy_db.pt"

  #14 opened 2 years ago by Ask-Yang
  3

• question about amass_parser.py

  #33 opened a year ago by dsdf783
  1

• Reproducing paper results

  #32 opened a year ago by Radu1999
  1

• Training for action recognition

  #30 opened a year ago by Radu1999
  0

• continuous issue about rendering the sample

  #26 opened a year ago by sygyq305
  3

• train in num_frames == -2

  #31 opened a year ago by sygyq305
  0

• two extra loss terms: mmd and hessian_penalty

  #28 opened a year ago by zshyang
  3

• results on HumanML3D dataset

  #29 opened a year ago by zshyang
  1

• SImilarities computed using motion and text embeddings are incorrect

  #27 opened a year ago by sohananisetty
  2

• use smplx add-on

  #24 opened a year ago by yinkangning0124
  3

• Why input size is 25 x 6?

  #19 opened a year ago by kjw9899
  6

• issue about rendering the sample

  #25 opened a year ago by sygyq305
  13

• The generated action is reversed

  #23 opened a year ago by sygyq305305
  3

• Text-to-Motion issue

  #21 opened a year ago by sygyq305305
  3

• AMASS Dataset issue

  #22 opened a year ago by sygyq305305
  0

• environment creation using conda fails

  #17 opened a year ago by gawsalyan
  1

• What is the script duration_finetunning.py used for?

  #16 opened a year ago by nomorecoding
  1

• the test of editting and interpolation error

  #10 opened 2 years ago by wang-zm18
  10

• The generated motions are static?

  #5 opened 2 years ago by LinghaoChan
  7

• the body of Text Motion render result is opposite

  #11 opened 2 years ago by Zessay
  5

• Execution issues

  #12 opened 2 years ago by chinnusai25
  1

• TypeError: __init__() missing 3 required positional arguments: 'njoints', 'nfeats', and 'num_classes'

  #2 opened 2 years ago by TechnicalAnt
  2

• Could we ignore global translation and only take joint rotation as input to MotionCLIP?

  #9 opened 2 years ago by mingdianliu
  1

• Two questions

  #8 opened 2 years ago by rbler1234
  2

• How to get the rendered image and the rigged character?

  #7 opened 2 years ago by yizhouzhao
  2

• Training translation??

  #6 opened 2 years ago by XiSHEN0220
  2

• About AMSS Dataset

  #3 opened 2 years ago by LinghaoChan
  3

• "action2motion_joints" list & AMASS db file does not include translation vector

  #4 opened 2 years ago by MenelauseZzw
  1

• code release?

  #1 opened 3 years ago by azadis
  8