antocad/FocusOnDepth

Issues

• training for relative depth

  #28 opened 2 months ago by YacineDeghaies
  0

• Training using SUNRGBD dataset

  #13 opened 2 years ago by sreejithpai
  1

• Question about gradient loss

  #27 opened 7 months ago by VincentItaliano
  0

• Thanks for a repo! Segmentation is wrong.

  #6 opened a year ago by onoyuki081
  9

• expected depth type ?

  #19 opened a year ago by iTomxy
  4

• vit_base_patch32_224 instead of vit_base_patch16_384

  #12 opened a year ago by 0xrushi
  1

• About the datasets

  #10 opened a year ago by shenbunny
  1

• changes to the patch embedding layer

  #9 opened a year ago by lidiatekeste2312
  2

• How to specify validation dataset

  #4 opened a year ago by caginozkaya
  1

• License

  #25 opened a year ago by jasapienza
  1

• No need for depth aspect

  #26 opened a year ago by SJayYang
  1

• Error Loading my own trained model.

  #24 opened a year ago by saqibnaziir
  0

• Unexpected key(s) in state_dict when type="depth"

  #14 opened 2 years ago by andy500
  7

• Inverse Depth

  #23 opened 2 years ago by 3togo
  0

• "compute_scale_and_shift" using median + MAD vs least square

  #22 opened 2 years ago by 3togo
  1

• Will FocusOnDepth support nn.DataParallel?

  #21 opened 2 years ago by 3togo
  0

• zipfile.BadZipFile: File is not a zip file

  #20 opened 2 years ago by 3togo
  1

• question in the focusdepth scripts

  #18 opened 2 years ago by visonpon
  0

• selected frames per dataset

  #16 opened 2 years ago by ahkwanzy
  0

• How to add IOU score?

  #8 opened 2 years ago by 0xrushi
  2

• about accuracy

  #11 opened 2 years ago by Gyeongnyeon-kim
  0

• Questions about Midas Training

  #3 opened 2 years ago by Only-bottle
  4

• Training loss and log

  #5 opened 2 years ago by jiaqixuac
  1

• Amazing!! Will the same weights work for "type":"segmentation" ?

  #7 opened 2 years ago by 0xrushi
  5

• A question about depth estimation accuracy .

  #2 opened 3 years ago by jiangchaokang
  2

• Hi. I have a question.

  #1 opened 3 years ago by Heloworkdl
  8