University of Limerick
Browse

HNAS: Hyper neural architecture search for image segmentation

Download (7.87 MB)
conference contribution
posted on 2022-11-29, 14:18 authored by Yassir Houreh, Mahsa MahdinejadMahsa Mahdinejad, ENRIQUE NAREDOENRIQUE NAREDO, Douglas Mota DiasDouglas Mota Dias, Conor RyanConor Ryan

Deep learning is a well suited approach to successfully address image processing and there are several Neural Networks architectures proposed on this research field, one interesting example is the U-net architecture and and its variants. This work proposes to automatically find the best architecture combination from a set of the current most relevant U-net architectures by using a genetic algorithm (GA) applied to solve the Retinal Blood Vessel Segmentation (RVS), which it is relevant to diagnose and cure blindness in diabetes patients. Interestingly, the experimental results show that avoiding human-bias in the design, GA finds novel combinations of U-net architectures, which at first sight seems to be complex but it turns out to be smaller, reaching competitive performance than the manually designed architectures and reducing considerably the computational effort to evolve them. 

Funding

SFI Centre for Research Training in Artificial Intelligence

Science Foundation Ireland

Find out more...

Automatic Design of Digital Circuits (ADDC)

Science Foundation Ireland

Find out more...

History

Publication

Proceedings of the 13th International Conference on Agents and Artificial Intelligence, 2, pp. 246-256

Publisher

ScitPress

Other Funding information

The fourth author is partially financed by the Coordenao de Aperfeioamento de Pessoal de Nvel Superior - Brasil (CAPES) - Finance Code 001

Also affiliated with

  • LERO - The Irish Software Research Centre

Department or School

  • Computer Science & Information Systems

Usage metrics

    University of Limerick

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC