figshare
Browse

Uncertainty-Guided Contrastive Learning For Single Source Domain Generalisation

Download (8.52 MB)
Version 3 2024-07-26, 14:09
Version 2 2024-01-04, 12:01
conference contribution
posted on 2024-07-26, 14:09 authored by Anastasios Arsenos, Dimitrios Kollias, Evangelos Petrongonas, Christos Skliros, Stefanos KolliasStefanos Kollias

A new model is presented in the paper for single source domain generalisation, through augmentation of input and label spaces and using contrastive learning. Uncertainty estimation is also generated at inference time. Experimental results illustrate the improved performance produced by the presented approach.

History

School affiliated with

  • School of Computer Science (Research Outputs)

Publication Title

ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Publisher

IEEE Xplore

External DOI

ISBN

979-8-3503-4486-8

eISBN

979-8-3503-4485-1

Date Submitted

2023-09-06

Date Accepted

2023-12-13

Date of Final Publication

2024-03-18

Event Name

2024 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2024),COEX, Seoul, Korea

Event Dates

14-19 April 2024

Open Access Status

  • Not Open Access

Date Document First Uploaded

2023-12-17

Publisher statement

© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Usage metrics

    University of Lincoln (Research Outputs)

    Categories

    Keywords

    domain generalisationaugmentationuncertainty estimationadversarial & contrastive learning

    Licence

    All Rights Reserved

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC