YvanYin/Metric3D

Issues

• How to improve the metric depth value?

  #161 opened 4 months ago by MoAbbasid
  9

• Ground truth for surface normal estimation

  #191 opened 22 days ago by Jiahao000
  0

• Weird sky, faraway object prediction when finetuning on nuScenes dataset

  #190 opened 22 days ago by Iven-Wu
  0

• If my dataset, which does not have camera intrinsics, can be used for fine-tuning the training of Metric3D?

  #189 opened 22 days ago by Victoire7
  0

• Running train.sh reports an error

  #188 opened 22 days ago by quezhiyu
  0

• Request for code of datasets preprocess

  #187 opened a month ago by zero0kiriyu
  1

• ViT Support For CUDA Versions >11.8

  #186 opened a month ago by Joeljensen28
  0

• How do we train HourglassDecoder?

  #184 opened a month ago by yutanpo1227
  1

• Great job! How do I get the normal labels for each frame from the scannetv2 data set

  #185 opened a month ago by li12138cmd
  0

• Normal Maps aren't generating

  #183 opened a month ago by Joeljensen28
  0

• 直角物体生成的点云图是斜的

  #182 opened a month ago by xiaomaxiao
  0

• When using torch.hub to do monocular estimation, it got RuntimeError

  #151 opened 4 months ago by LucioSunj
  7

• Please ask the code how to use eth3d dataset, I see that there is eth3d_dataset.py, how to run it

  #181 opened 2 months ago by wudabingm
  0

• How about outputting the depth value to a .txt file, what code do I need to add?

  #180 opened 2 months ago by wudabingm
  0

• Can't run the test_vit.sh

  #172 opened 3 months ago by CDchenlin
  3

• Converting Depth Map Pixel Values to Real-World Units (Meters)

  #169 opened 2 months ago by Moeez-Muslim
  1

• how can I get these pretrained ckpts of droid-slam

  #179 opened 2 months ago by ziyue-YING
  0

• How to achieve metrics as in the paper?

  #178 opened 2 months ago by WLatonov
  0

• Loss when training with my own data

  #177 opened 2 months ago by Sissi-lu
  0

• how to run hubconf。

  #176 opened 2 months ago by gdy160607
  0

• How to train from scratch

  #175 opened 2 months ago by Linengyao
  0

• Running inference on CPU

  #149 opened 4 months ago by AD-lite24
  4

• training error

  #160 opened 2 months ago by stan1512
  1

• Inquiry About 3D Reconstruction Code Availability

  #174 opened 2 months ago by CaiSir0075
  2

• Get a depth dataset from REPLICA dataset

  #171 opened 3 months ago by czeyveli1
  1

• Questions on fine-tuning freeze and config parameters

  #173 opened 3 months ago by tazalapizza
  1

• about model

  #167 opened 3 months ago by rimu123
  1

• do_test.py casting data_basic as both a Dict and a Namespace

  #165 opened 3 months ago by spomichter
  1

• Encountering an issue with the conversion to CoreML.

  #164 opened 3 months ago by 1e-cha4-end0rmi
  1

• Any plan for releasing all the training dataset?

  #157 opened 4 months ago by inkzk
  1

• 目前的代码只能在GPU:0中运行吗？

  #159 opened 4 months ago by sukasu403
  1

• 关于相机传感器不影响绝对深度估计

  #168 opened 3 months ago by zhjw0927
  1

• Request for Surface Normal Code

  #163 opened 3 months ago by CDchenlin
  1

• Licensing question

  #170 opened 3 months ago by spiegelball
  1

• eval on NYU dataset

  #154 opened 4 months ago by shawnricecake
  4

• Prediction shape mismatch with GroundTruth

  #156 opened 4 months ago by saadehmd
  13

• Dense depth map output

  #166 opened 3 months ago by Deng-King
  1

• Focal length in mm or pixels?

  #162 opened 4 months ago by spomichter
  1

• Metric depth recovery part

  #153 opened 4 months ago by xyclone10
  2

• Infinite Depth

  #145 opened 5 months ago by HardikJain02
  3

• DType bfloat16 issue

  #158 opened 4 months ago by robofar
  1

• Dataset content issues

  #155 opened 4 months ago by oywenjun11
  2

• Disabling the normal estimation during inference

  #146 opened 4 months ago by basketballmypassion
  1

• Which one is better, metric3dv2 or geowizard?

  #152 opened 4 months ago by StarsTesla
  1

• What is the python version used to run a stable version of the project

  #150 opened 4 months ago by AD-lite24
  1

• How to achieve corresponding 2D points for each 3D points?

  #142 opened 5 months ago by HardikJain02
  2

• Code for camera intrinsic extraction from EXIF

  #148 opened 4 months ago by kam1107
  0

• Why it has no reaction on classroom pictures?

  #147 opened 5 months ago by lovehuanhuan
  2

• Expected onnx model input mismatch with name in test_onnx.py

  #143 opened 5 months ago by norbertlink
  1

• Are the VIT checkpoints on Huggingface the finetuned ones on NYU?

  #144 opened 5 months ago by Guirassy43
  1