Computer-aided analysis of biological images typically requires extensive training on large-scale annotated datasets, which is not viable in many situations. In this paper, we present Generative Adversarial Network Discriminator Learner (GAN-DL), a novel self-supervised learning paradigm based on the StyleGAN2 architecture, which we employ for self-supervised image representation learning in the case of fluorescent biological images.

Exploiting generative self-supervised learning for the assessment of biological images with lack of annotations / Mascolini, Alessio; Cardamone, Dario; Ponzio, Francesco; Di Cataldo, Santa; Ficarra, Elisa. - In: BMC BIOINFORMATICS. - ISSN 1471-2105. - 23:1(2022), pp. 295-312. [10.1186/s12859-022-04845-1]

Exploiting generative self-supervised learning for the assessment of biological images with lack of annotations

Ficarra, Elisa
Supervision
2022

Abstract

Computer-aided analysis of biological images typically requires extensive training on large-scale annotated datasets, which is not viable in many situations. In this paper, we present Generative Adversarial Network Discriminator Learner (GAN-DL), a novel self-supervised learning paradigm based on the StyleGAN2 architecture, which we employ for self-supervised image representation learning in the case of fluorescent biological images.
2022
24-lug-2022
23
1
295
312
Exploiting generative self-supervised learning for the assessment of biological images with lack of annotations / Mascolini, Alessio; Cardamone, Dario; Ponzio, Francesco; Di Cataldo, Santa; Ficarra, Elisa. - In: BMC BIOINFORMATICS. - ISSN 1471-2105. - 23:1(2022), pp. 295-312. [10.1186/s12859-022-04845-1]
Mascolini, Alessio; Cardamone, Dario; Ponzio, Francesco; Di Cataldo, Santa; Ficarra, Elisa
File in questo prodotto:
File Dimensione Formato  
Mascolini_GAN_BMC-Bioinformatics_2022.pdf

Open access

Tipologia: Versione pubblicata dall'editore
Dimensione 2.74 MB
Formato Adobe PDF
2.74 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

Licenza Creative Commons
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11380/1284304
Citazioni
  • ???jsp.display-item.citation.pmc??? 5
  • Scopus 13
  • ???jsp.display-item.citation.isi??? 10
social impact