gmberton/CosPlace

Issues

• where are intrinsics and exintrinsics?

  #48 opened 3 months ago by 9796l
  1

• About the training time

  #4 opened 2 years ago by Orbis36
  2

• Issues about dataset download speed

  #47 opened 8 months ago by dl857
  2

• Camera Intrinsic Matrix of SF-XL dataset

  #46 opened 9 months ago by LMN1590
  1

• do image retrieval on my own dataset

  #45 opened 10 months ago by xuxumiao777
  2

• Make custom dataset of other city alternative to SF-XL

  #43 opened 10 months ago by songkemail
  2

• Can't download small version of dataset...

  #44 opened 10 months ago by rocker12121
  3

• Validation Logs

  #42 opened 10 months ago by Shai2022
  8

• Questions about performance during training

  #41 opened a year ago by Devoe-97
  4

• Question about the 'groups_num' for training.

  #40 opened a year ago by Devoe-97
  2

• Peformance of Resnet18_512.pth

  #38 opened a year ago by Cc19245
  2

• License of trained weights?

  #35 opened a year ago by Parskatt
  2

• Method for Downloading Historical Google Street View Images

  #34 opened a year ago by AuderySun
  1

• ViT trained models

  #33 opened 2 years ago by oravus
  4

• Why is iter used here during training iterations instead of directly using the dataloader?

  #32 opened 2 years ago by wpumain
  8

• Why do I only perform Data Augmentation in CPU mode?

  #31 opened 2 years ago by wpumain
  2

• Image rotation in augmentations

  #30 opened 2 years ago by Danessely
  2

• About using the model in other datasets

  #29 opened 2 years ago by Danessely
  7

• How to add orientation to 360 degree panorama photo?

  #26 opened 2 years ago by wpumain
  6

• Can you tell me what the name of the image file from the 【raw/train/panoramas】 data means?

  #27 opened 2 years ago by wpumain
  5

• How are these SF-XL downloaded from Google Earth?

  #28 opened 2 years ago by wpumain
  2

• Error when trying to use own dataset

  #24 opened 2 years ago by MysteryHS
  4

• Where can I get Pitts250k data?

  #25 opened 2 years ago by wpumain
  2

• What else do you think CosPlace can improve?

  #23 opened 2 years ago by wpumain
  2

• In the training set, the content of the picture is a house far away from the photographer, so will it cause a big error to use a close-range picture of the house to predict in test time?

  #18 opened 2 years ago by wpumain
  8

• Can you tell me how long it takes to train on about 900G raw data ?

  #20 opened 2 years ago by wpumain
  7

• When the model makes predictions, what is the relationship between the number of images in the query dataset and the number of database images?

  #22 opened 2 years ago by wpumain
  2

• Why only freeze layers before conv_3 of ResNet？

  #21 opened 2 years ago by wpumain
  2

• Can the evaluation of this model be used in other city images?

  #19 opened 2 years ago by wpumain
  4

• Is the UTM coordinate of the picture in the SF-XL dataset the coordinate of the camera when taking the picture?

  #17 opened 2 years ago by wpumain
  2

• About evaluation in SF-XL

  #15 opened 2 years ago by ShihaoShao-GH
  6

• Question about SARE and SFRS results (tested on tokyo247 dataset) used in the paper

  #12 opened 2 years ago by hit2sjtu
  3

• HDD may cause training to be too slow.

  #14 opened 2 years ago by BinuxLiu
  2

• About evaluation on the other datasets

  #13 opened 2 years ago by YellowBe
  2

• using in remote sensing visual place recognition

  #9 opened 2 years ago by rafal7466
  1

• using multi-GPU for training step

  #10 opened 2 years ago by rafal7466
  1

• Performance Comparison with NetVLAD

  #8 opened 2 years ago by divyagupta25
  3

• Question about coordinate system

  #7 opened 2 years ago by YangSooRim
  1

• Question about testing image size.

  #6 opened 2 years ago by hao-qiang
  1

• Question about the number of 'group'

  #5 opened 2 years ago by hmf21
  1

• Cannot open SF-XL datasat

  #3 opened 3 years ago by seawavve
  2

• SF-XL

  #2 opened 3 years ago by xuzhuang1996
  1

• About the training dataset

  #1 opened 3 years ago by Barry-Liang
  4