Despite being very powerful in standard learning settings, deep learning models can be extremely brittle when deployed in scenarios different from those on which they were trained. Domain generalization methods investigate this problem and data augmentation strategies have shown to be helpful tools to increase data variability, supporting model robustness across domains. In our work we focus on style transfer data augmentation and we present how it can be implemented with a simple and inexpensive strategy to improve generalization. Moreover, we analyze the behavior of current state of the art domain generalization methods when integrated with this augmentation solution: our thorough experimental evaluation shows that their original effect almost always disappears with respect to the augmented baseline. This issue open new scenarios for domain generalization research, highlighting the need of novel methods properly able to take advantage of the introduced data variability.

Rethinking Domain Generalization Baselines / Cappio, F; D'Innocente, A; Tommasi, T. - (2021), pp. 9227-9233. (Intervento presentato al convegno 25th International Conference on Pattern Recognition, ICPR 2020 tenutosi a Milan; Italy) [10.1109/ICPR48806.2021.9412735].

Rethinking Domain Generalization Baselines

D'Innocente A
Secondo
;
Tommasi T
Ultimo
2021

Abstract

Despite being very powerful in standard learning settings, deep learning models can be extremely brittle when deployed in scenarios different from those on which they were trained. Domain generalization methods investigate this problem and data augmentation strategies have shown to be helpful tools to increase data variability, supporting model robustness across domains. In our work we focus on style transfer data augmentation and we present how it can be implemented with a simple and inexpensive strategy to improve generalization. Moreover, we analyze the behavior of current state of the art domain generalization methods when integrated with this augmentation solution: our thorough experimental evaluation shows that their original effect almost always disappears with respect to the augmented baseline. This issue open new scenarios for domain generalization research, highlighting the need of novel methods properly able to take advantage of the introduced data variability.
2021
25th International Conference on Pattern Recognition, ICPR 2020
Computer Science; Computer Vision and Pattern Recognition; Computer Science - Machine Learning
04 Pubblicazione in atti di convegno::04b Atto di convegno in volume
Rethinking Domain Generalization Baselines / Cappio, F; D'Innocente, A; Tommasi, T. - (2021), pp. 9227-9233. (Intervento presentato al convegno 25th International Conference on Pattern Recognition, ICPR 2020 tenutosi a Milan; Italy) [10.1109/ICPR48806.2021.9412735].
File allegati a questo prodotto
File Dimensione Formato  
CoppioBorlino_Rethinking_2021.pdf

accesso aperto

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Creative commons
Dimensione 2.02 MB
Formato Adobe PDF
2.02 MB Adobe PDF

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11573/1504633
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 16
  • ???jsp.display-item.citation.isi??? 9
social impact