amaralibey/Bag-of-Queries

Issues

• About Nordland* and Nordland** results for DINOV2

  #11 opened 3 months ago by xjh19971
  2

• About the version of Nordland!

  #9 opened a month ago by STSTERANDMOMO
  1

• How to use multi-gpu training?

  #12 opened 3 months ago by sumitmishra209
  1

• impact of batch size

  #13 opened 3 months ago by sumitmishra209
  1

• Minimal example of inference

  #14 opened a month ago by tianyilim
  2

• Help Needed to Regenerate Attention Heatmaps for VPR Project

  #15 opened a month ago by akselpoulsen
  1

• For more details about the learning rate

  #10 opened 3 months ago by Li-Yun-star
  5

• About DinoV2+BoQ Training.

  #8 opened 5 months ago by BinuxLiu
  3

• Feature dimensions of BoQ wth DinoV2 backbone

  #5 opened 6 months ago by BinuxLiu
  14

• About DinoV2+BoQ training and released checkpoint

  #7 opened 6 months ago by hit2sjtu
  3

• The Hyper-parameters for training the DinoV2+BoQ?

  #6 opened 6 months ago by Frost24K
  2

• Great work! I want to konw when using DINOv2 as the backbone, what are the training image size and the testing image size, respectively? And what is the accuracy on the MSLS challenge? Looking forward to your reply!

  #3 opened 6 months ago by STSTERANDMOMO
  7

• Nordland*

  #4 opened 6 months ago by LKELN
  2

• When to release the code and the pretrained model?

  #1 opened 6 months ago by Frost24K
  1

• Your research results are impressive! I'm curious about how to implement the "learned Bag of Queries" mentioned in the paper. After reading the paper, it seems to be related to the size of the batch_size. How can the reproducibility of the results be ensured? I hope you can publish your code soon, as I am very interested! Thank you!

  #2 opened 6 months ago by STSTERANDMOMO
  1