prs-eth/Marigold

Issues

• How to calculate depth in Meters from the depth map in numpy

  #123 opened 3 months ago by Luqman-Ansari
  4

• Depth Map Prior

  #137 opened 20 days ago by sam-thomas-dd
  3

• About training loss

  #136 opened a month ago by whyandbecause
  2

• Fine Tuning on Custom Dataset

  #135 opened a month ago by Luqman-Ansari
  1

• Is it possible to improve the model's performance by switching the base model to SDXL?

  #132 opened a month ago by kke19
  1

• VAE on depth maps

  #134 opened a month ago by 2hiTee
  1

• Running on cpu only ?

  #133 opened 2 months ago by ThomasCartier
  1

• Request for Training Code of Normal Generation Model

  #131 opened 2 months ago by kke19
  0

• The training runs abnormally, and the memory is occupied at a certain step

  #129 opened 3 months ago by zad37
  5

• Other training metric are abnormal

  #130 opened 2 months ago by zad37
  2

• loss=nan occurs during training

  #111 opened 4 months ago by dongyyyy
  5

• The installation error

  #127 opened 2 months ago by WLatonov
  1

• Trying to resize storage that is not resizable

  #117 opened 2 months ago by dongyyyy
  1

• Can Marigold be trained from scratch?

  #126 opened 2 months ago by resurgo97
  1

• Inquiry About 3D Reconstruction Code Availability

  #128 opened 3 months ago by CaiSir0075
  1

• About some problems that arise during training

  #95 opened 5 months ago by water221
  16

• VKitti dataset setup

  #125 opened 3 months ago by jameelhassan
  1

• Why is the inferencing process currently stuck and not making any progress

  #124 opened 3 months ago by Abbsalehi
  1

• Rotation of the point cloud

  #120 opened 4 months ago by fkjkey
  1

• Where is the core code for 'duplicate the weight tensor of the input layer'?

  #118 opened 4 months ago by Kyfafyd
  1

• conflict in .yaml file

  #121 opened 4 months ago by Meghdad24
  2

• Numpy is not available error

  #119 opened 4 months ago by GLtangzero
  0

• Question Regarding Poor Performance on ScanNet After Code Reproduction

  #116 opened 4 months ago by haopengmai
  3

• NYUv2 Filled depth map

  #114 opened 5 months ago by gaoxiangjun
  2

• Wandb logging is always interrupted unexpectedly

  #115 opened 5 months ago by water221
  1

• Fine-tune the base model

  #101 opened 5 months ago by meiqing-wang
  1

• Muti-GPUs training

  #98 opened 5 months ago by Smarome
  1

• Changed loss function to mean_abs_rel for scale-invariant depth but training is unstable

  #112 opened 5 months ago by YacineDeghaies
  1

• Release of LCM drop?

  #113 opened 5 months ago by omrastogi
  0

• Unusual slow training speed

  #94 opened 6 months ago by Fra-Tsuna
  9

• noisy inference, any idea why ?

  #106 opened 5 months ago by YacineDeghaies
  5

• Questions about LCM model training

  #103 opened 5 months ago by luochunfen
  3

• Validate using data with different resolutions

  #109 opened 5 months ago by dongyyyy
  1

• Use data sets with different resolutions for training

  #108 opened 5 months ago by dongyyyy
  1

• Use data sets of different resolutions for training

  #107 opened 5 months ago by dongyyyy
  0

• Data cannot be read during training

  #99 opened 5 months ago by dongyyyy
  1

• Validation Error

  #100 opened 5 months ago by Keivin98
  1

• noob questions about code understanding.

  #104 opened 5 months ago by YacineDeghaies
  3

• [HuggingFace] is it possible to fine-tune Marigold using the AutoTrainer on my custom dataset ?

  #105 opened 5 months ago by YacineDeghaies
  0

• What does the consistent value of 721.5377 represent in eigen_test_files_with_gt.txt?

  #102 opened 5 months ago by haopengmai
  1

• Is it possible to train the model by feeding it rgb images + 8-bit relative depth maps ?

  #97 opened 5 months ago by YacineDeghaies
  1

• issue with reproducing results in the paper

  #96 opened 6 months ago by yongmayer
  1

• How to organize the vkitti data

  #91 opened 6 months ago by Larissa0829
  1

• do you plan to release better and more accurate models for this original marigold?

  #87 opened 7 months ago by mikebilly
  1

• Any reason for not using vae.std to generate RGB latent?

  #86 opened 6 months ago by Jerrypiglet
  1

• Clarification Needed: Training and Inference Pipeline

  #85 opened 6 months ago by ShenZheng2000
  4

• train the model on my custom dataset

  #92 opened 6 months ago by YacineDeghaies
  1

• The demo 3D looks ok but not match with the predicted depth image

  #90 opened 6 months ago by N-NEJATISHAHIDIN
  1

• where can I get the "output/marigold_base/checkpoint/latest"

  #89 opened 6 months ago by Larissa0829
  1

• The purpose of using v_prediction as the target?

  #88 opened 6 months ago by yongmayer
  1