EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Point-SPV: end-to-end enhancement of object recognition in simulated prosthetic vision using synthetic viewing points

Ashkan Nejad, Burcu Küçükoǧlu, Jaap de Ruyter van Steveninck, Sandra Bedrossian, Joost Heutink, Gera A de Haan, Frans W Cornelissen, Marcel van Gerven

Frontiers in Human Neuroscience · 24.03.2025 · 19

DOI: 10.3389/fnhum.2025.1549698 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Human Neuroscience, online erschienen am 24.03.2025. Verfasst von Ashkan Nejad und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Research and Improvement of Care, Royal Dutch Visio, Huizen, Netherlands. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Horizon 2020, Horizon 2020 Framework Programme.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Prosthetic vision systems aim to restore functional sight for visually impaired individuals by replicating visual perception by inducing phosphenes through electrical stimulation in the visual cortex, yet there remain challenges in visual representation strategies such as including gaze information and task-dependent optimization. In this paper, we introduce Point-SPV, an end-to-end deep learning model designed to enhance object recognition in simulated prosthetic vision. Point-SPV takes an initial step toward gaze-based optimization by simulating viewing points, representing potential gaze locations, and training the model on patches surrounding these points. Our approach prioritizes task-oriented representation, aligning visual outputs with object recognition needs. A behavioral gaze-contingent object discrimination experiment demonstrated that Point-SPV outperformed a conventional edge detection method, by facilitating observers to gain a higher recognition accuracy, faster reaction times, and a more efficient visual exploration. Our work highlights how task-specific optimization may enhance representations in prosthetic vision, offering a foundation for future exploration and application.

Schlagwörter

Stichwörter der Autor:innen

deep learningend-to-end trainingobject recognitionsimulated prosthetic visionsynthetic viewing points

Autor:innen und Institutionen

  • Department of Research and Improvement of Care, Royal Dutch Visio, Huizen, Netherlands.
  • Department of Machine Learning and Neural Computing, Donders Institute for Brain, Cognition and Behaviour, Radboud University, Nijmegen, Netherlands.
  • Laboratory for Experimental Ophthalmology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands.
  • Faculty of Science and Engineering, University of Groningen, Groningen, Netherlands.
  • Department of Clinical and Developmental Neuropsychology, University of Groningen, Groningen, Netherlands.

Förderung

  • Horizon 2020
  • Horizon 2020 Framework Programme

Zitieren

APA

Nejad, A., Küçükoǧlu, B., de Ruyter van Steveninck, J., Bedrossian, S., Heutink, J., de Haan, G. A., Cornelissen, F. W., & van Gerven, M. (2025). Point-SPV: end-to-end enhancement of object recognition in simulated prosthetic vision using synthetic viewing points. Frontiers in Human Neuroscience, 19. https://doi.org/10.3389/fnhum.2025.1549698

Vancouver

Nejad A, Küçükoǧlu B, de Ruyter van Steveninck J, Bedrossian S, Heutink J, de Haan GA, et al. Point-SPV: end-to-end enhancement of object recognition in simulated prosthetic vision using synthetic viewing points. Frontiers in Human Neuroscience. 2025;19. doi:10.3389/fnhum.2025.1549698

BibTeX
@article{Nejad2025,
  author = {Nejad, Ashkan and Küçükoǧlu, Burcu and de Ruyter van Steveninck, Jaap and Bedrossian, Sandra and Heutink, Joost and de Haan, Gera A and Cornelissen, Frans W and van Gerven, Marcel},
  title = {Point-SPV: end-to-end enhancement of object recognition in simulated prosthetic vision using synthetic viewing points},
  journal = {Frontiers in Human Neuroscience},
  year = {2025},
  volume = {19},
  doi = {10.3389/fnhum.2025.1549698}
}
Kostenlos starten →