Parskatt/DKM

Issues

• Generate custom dataset

  #65 opened 3 months ago by nfyfamr
  2

• Error occurs if not GPU:0

  #66 opened 3 months ago by randomwons
  0

• Unexpected Match Point with Transparent Background

  #63 opened 5 months ago by qidihan
  5

• about image resolution

  #64 opened 5 months ago by KN-Zhang
  1

• Keep aspect ratio of images?

  #62 opened 5 months ago by javiabellan
  1

• Training on Kitti

  #61 opened 5 months ago by StardustLu
  0

• Inference in batched mode: all values in certainty is zeros

  #59 opened 6 months ago by lllabmaster
  4

• Question about Coordinate Embeddings implementations

  #58 opened 7 months ago by xjiangan
  5

• DKMv3 and DKMv2

  #41 opened a year ago by sjtuztq
  4

• how to export a onnx model?

  #57 opened 7 months ago by bfdong
  0

• How to use GC-RANSAC for pose estimation?

  #47 opened a year ago by GavinYang5
  1

• question about internal resolution and homography estimation

  #56 opened 7 months ago by jlzhou1
  0

• scannet test code

  #54 opened 8 months ago by jdk9405
  1

• if it's possible to use DKMv3 for training when there is no depth information in my own dataset?

  #49 opened a year ago by DiKu-nn
  3

• How to draw the Figure 4. Coordinate embeddings preserve multimodality in the main paper

  #53 opened 8 months ago by HustBestCat
  2

• A small detail question about dense-flow

  #52 opened 8 months ago by chen9run
  1

• The issues encountered during training the network.

  #51 opened 8 months ago by stilcrad
  1

• Whether the inference speed can be optimized

  #50 opened 8 months ago by NickJony
  1

• torch.linalg.inv

  #45 opened a year ago by wtishere
  6

• question about loading the training images

  #48 opened a year ago by Master-cai
  2

• synthetic dataset

  #46 opened a year ago by lyp19
  0

• About previous sota pdc-net+

  #44 opened a year ago by Moreland-cas
  5

• np.random.choice for good_samples

  #4 opened 2 years ago by ufukefe
  10

• Questions about the key points on Megadepth test images

  #43 opened a year ago by xmlyqing00
  2

• Pretrained weights licensing

  #42 opened a year ago by stepankonev
  2

• About testing results

  #40 opened a year ago by sugar-fly
  2

• When using multi-GPU training, there is additional memory occupancy on GPU 0

  #38 opened a year ago by qkqhd222
  1

• About the pretrained model with resnet18

  #37 opened a year ago by longmao-yiran
  1

• Questions about global matcher

  #36 opened a year ago by hanquansanren
  3

• About the pretrained model

  #34 opened a year ago by sugar-fly
  3

• e_R reaches 180°

  #33 opened a year ago by Moreland-cas
  15

• what does low_res_certainty do in dkm.py?

  #31 opened a year ago by Moreland-cas
  3

• the structure and information of 'warp'

  #29 opened 2 years ago by zzxq2310
  5

• run on batch data

  #28 opened 2 years ago by whulc
  5

• `numpy.random.choice` to pytorch?

  #27 opened 2 years ago by ducha-aiki
  6

• Inference time

  #20 opened 2 years ago by ckLibra
  2

• Can DKM run on CPU only?

  #21 opened 2 years ago by ducha-aiki
  9

• 3d point projection, best way to fetch matches

  #23 opened 2 years ago by drperpen
  6

• How many GPUs and consumed memory for your training?

  #19 opened 2 years ago by TruongKhang
  7

• DKM:sample

  #18 opened 2 years ago by scott-vsi
  2

• purpose of conf_thresh parameter in sample function is confusing

  #11 opened 2 years ago by Tobimaru
  3

• It is fair to compare DKM directly with LoFTR when DKM uses more training data sampled from MegaDepth ?

  #15 opened 2 years ago by noone-code
  15

• Did you test your DKM in the Aachen Benchmark ?

  #14 opened 2 years ago by noone-code
  3

• when demo dkmV3 release ?

  #13 opened 2 years ago by QuanNguyen94
  2

• DKMv2 and DKMv3 paper

  #12 opened 2 years ago by chicleee
  0

• running demo_match.py without GPUs

  #9 opened 2 years ago by cxmscb
  1

• Does the DKM requires the fixed size (384*512) images as input ?

  #10 opened 2 years ago by noone-code
  2

• Match Score

  #8 opened 2 years ago by atharvjairath
  4

• With Synthetic Dataset Training Codes

  #7 opened 2 years ago by TWJianNuo
  6

• Logics in function "upsample_preds" hard to understand

  #5 opened 2 years ago by TWJianNuo
  2