Accepted author manuscript, 2.4 MB, PDF document
Available under license: CC BY: Creative Commons Attribution 4.0 International License
Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review
Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review
}
TY - CONF
T1 - Unsupervised Domain Adaptation within Deep Foundation Latent Spaces
AU - Kangin, Dmitry
AU - Angelov, Plamen
PY - 2024/5/11
Y1 - 2024/5/11
N2 - The vision transformer-based foundation models, such as ViT or Dino-V2, are aimed at solving problems with little or no finetuning of features. Using a setting of prototypical networks, we analyse to what extent such foundation models can solve unsupervised domain adaptation without finetuning over the source or target domain. Through quantitative analysis, as well as qualitative interpretations of decision making, we demonstrate that the suggested method can improve upon existing baselines, as well as showcase the limitations of such approach yet to be solved. The code is available at: https://github.com/lira-centre/vit_uda/
AB - The vision transformer-based foundation models, such as ViT or Dino-V2, are aimed at solving problems with little or no finetuning of features. Using a setting of prototypical networks, we analyse to what extent such foundation models can solve unsupervised domain adaptation without finetuning over the source or target domain. Through quantitative analysis, as well as qualitative interpretations of decision making, we demonstrate that the suggested method can improve upon existing baselines, as well as showcase the limitations of such approach yet to be solved. The code is available at: https://github.com/lira-centre/vit_uda/
M3 - Conference paper
T2 - ICLR 2024 2nd Workshop on Mathematical and Empirical Understanding of Foundation Models (ME-FoMo)
Y2 - 7 May 2024 through 11 May 2024
ER -