Skip to content

CercaTrovato/MVSIB-pro

Repository files navigation

MVSIB-pro

This paper introduces a unified MVC model (MVSIB). The model adaptively differentiates false negatives from hard negatives through uncertainty measurement and curriculum learning, while enhancing feature discrimination with a combination of soft masking and adversarial loss.

About

This paper introduces a improved MVSIB model (MVSIB-pro). The model adaptively differentiates false negatives from hard negatives through uncertainty measurement and curriculum learning, while enhancing feature discrimination with a combination of soft masking and adversarial loss.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors