Skip to content

Implemented a Simulation-to-Reality transfer learning strategy using ResNet18 and adversarial discriminative domain adaptation (ADDA) for wireless gesture recognition. This approach improved real-world dataset accuracy from 83.0% to 96.5%.

Notifications You must be signed in to change notification settings

rzy0901/CASTER_classification

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CASTER_classification

Source Only + Source mixed target w/ labels

  • Testing accuracies for Schemes 1~6:

accuracy

  • Confusion charts for Schemes 1~6:
Scheme 1 Scheme 2 Scheme 3
scheme 1 scheme 2 scheme 3
Scheme 4 Scheme 5 Scheme 6
scheme 4 scheme 5 scheme 6

Unsupervised domain adaptation

Source only (features.ipynb: 83%) Transferred (ADDA_test_new_5x10.ipynb: 96.5%)
scheme 1 TSNE

About

Implemented a Simulation-to-Reality transfer learning strategy using ResNet18 and adversarial discriminative domain adaptation (ADDA) for wireless gesture recognition. This approach improved real-world dataset accuracy from 83.0% to 96.5%.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published