wl-zhao/VPD

Issues

• Submodule update failed

  #67 opened a month ago by sinhasagar507
  1

• GPUs resource requirements

  #68 opened 10 days ago by caoer01
  0

• incompatible MMCV version of VPD seg

  #65 opened 2 months ago by FreeWillThorn
  1

• class embeddings

  #56 opened a year ago by baihetun
  1

• class_embedding of VOC PASCAL2012

  #66 opened a month ago by caoer01
  0

• Download pre-trained checkpoints

  #64 opened 2 months ago by caoer01
  0

• Consultation on mmcv and mmseg versions in the environment

  #60 opened 10 months ago by Shijie-Xiao
  2

• How to remove text information in denoising UNet based on the existing code?

  #63 opened 7 months ago by RuiTianHIT
  1

• miou50%，macc50%

  #62 opened 7 months ago by Saillxl
  2

• from mmcv.cnn.utils import revert_sync_batchnormfrom error

  #46 opened a year ago by 1103167965
  2

• VPDSeg model state dictionary keys missing

  #61 opened 7 months ago by apenzko
  1

• resolution

  #26 opened a year ago by elhamAm
  2

• RTX 4090Ti GPUs, Batch Size Reduction, and Performance Challenges in Depth Estimation

  #59 opened 10 months ago by wonjsu
  0

• How to use VPD for image translation

  #58 opened a year ago by rainsoulsrx
  0

• About cross attention in depth estimation

  #41 opened a year ago by minhohihi
  1

• Reproducing the performance of Referring Image Segmentation task

  #57 opened a year ago by golddohyun
  0

• UNetWrapper

  #55 opened a year ago by Maram-Helmy
  0

• No CLIP's preformance

  #54 opened a year ago by wangjiyuan9
  0

• error running depth inference on NYUv2

  #39 opened a year ago by kalyani7195
  3

• How can we run vpd for datasets like diode indoor or hypersim where no class ids are available

  #53 opened a year ago by k-styles
  0

• About dump_nyu_text_embeddings.py

  #52 opened a year ago by sym330
  0

• Is it possible to use VPD for detection task?

  #51 opened a year ago by WWW2323
  0

• Inference on NYU Depth doesnt work

  #50 opened a year ago by CanKeles5
  0

• Can you release the training logs on Depth estimation on NYUv2?

  #49 opened a year ago by develop-productivity
  3

• datasets can't be download

  #48 opened a year ago by kaidi00
  1

• Blurred depth maps

  #45 opened a year ago by Aradhye2002
  1

• NYUv2 datasets

  #47 opened a year ago by develop-productivity
  0

• How do you handle the info loss from VAE encoding?

  #44 opened a year ago by Darkbblue
  0

• Can you release the training log for segmentation?

  #43 opened a year ago by WasedaMagina
  1

• will you release the code of diffswap?

  #38 opened a year ago by vicentowang
  3

• how to inference single image using VPD？？

  #42 opened a year ago by dreamlychina
  0

• Failed to reproduce the accuracy on ADE20k

  #32 opened a year ago by Ben-Louis
  2

• Is using test file name for inference a fair practice?

  #34 opened a year ago by Aradhye2002
  2

• Question about "S"

  #40 opened a year ago by BlingHe
  1

• Training Time

  #29 opened a year ago by zgzxy001
  1

• error when using xformers

  #37 opened a year ago by baoxiaoyi
  3

• class embedding

  #35 opened a year ago by elhamAm
  1

• FPN head

  #36 opened a year ago by elhamAm
  1

• Reproducing 4k/8k segmentation results

  #30 opened a year ago by manuelknott
  3

• Mismatch between reported and observed performance of pretrained models made available

  #31 opened a year ago by Aradhye2002
  0

• The class_id seems to be incorrect in depth estimation

  #33 opened a year ago by Ben-Louis
  0

• t set to 0 in unet

  #28 opened a year ago by elhamAm
  1

• Reproduce the performance of ADE20K

  #23 opened a year ago by Yi-Qi638
  7

• Question around number of denoising steps

  #27 opened a year ago by agneet42
  2

• Loss calculation error during training

  #25 opened a year ago by agneet42
  1

• Output Dim of Depth Estimation

  #24 opened a year ago by agneet42
  1

• Struggling to setup VPD

  #21 opened a year ago by Sivloc
  3

• About class_embeddings

  #22 opened a year ago by NorthSummer
  2

• mm versions for segmentation

  #19 opened 2 years ago by nkondapa
  1

• typo in the paper

  #20 opened 2 years ago by sleekEagle
  1