Skip to content

Weakly Supervised Contrastive Learning for Vehicle Re-id

Notifications You must be signed in to change notification settings

andrewyu90/WSCL_VeReid

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

WSCL-SDA

Source code for the cvpr paper-5482 titled 'Contrastive leanring and domain adaptation using weak supervision for unsupervised vehicle Re-identification'

#Prepare dataset

  1. unzip dataset files into './data' directory
  2. change the ditectory names to 'bounding_box_train' (training set), 'bounding_box_test' (test set), 'query' (query set).
  • You may need reorganise the datasets.

#Train run train.sh

#Test run test.sh

#Reproduct experimental results Since the limitation of upload file size, we could not include the model files, So, to reproduct the experimental results, please follows bellow guide.

  1. Download learnt models frrom 'https://www.dropbox.com/s/256s0g5xtm4llk0/saved_models.zip?dl=0'
  2. unzip
  3. Run test.sh

About

Weakly Supervised Contrastive Learning for Vehicle Re-id

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published