EUVIMEDEUVIMED entdecken
Open Access

Predicting Goal-directed Attention Control Using Inverse-Reinforcement Learning

Gregory J Zelinsky, Yupei Chen, Seoyoung Ahn, Hossein Adeli, Zhibo Yang, Lihan Huang, Dimitrios Samaras, Minh Hoai

Neurons, Behavior, Data analysis, and Theory · 2021

DOI: 10.51628/001c.22322 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neurons, Behavior, Data analysis, and Theory, 2021. Verfasst von Gregory J Zelinsky und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Psychology, Stony Brook University, Stony Brook, NY, 11794, USA. Die Arbeit wurde bisher 10-mal zitiert (Crossref). Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NEI NIH HHS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Understanding how goals control behavior is a question ripe for interrogation by new methods from machine learning. These methods require large and labeled datasets to train models. To annotate a large-scale image dataset with observed search fixations, we collected 16,184 fixations from people searching for either microwaves or clocks in a dataset of 4,366 images (MS-COCO). We then used this behaviorally-annotated dataset and the machine learning method of inverse-reinforcement learning (IRL) to learn target-specific reward functions and policies for these two target goals. Finally, we used these learned policies to predict the fixations of 60 new behavioral searchers (clock = 30, microwave = 30) in a disjoint test dataset of kitchen scenes depicting both a microwave and a clock (thus controlling for differences in low-level image contrast). We found that the IRL model predicted behavioral search efficiency and fixation-density maps using multiple metrics. Moreover, reward maps from the IRL model revealed target-specific patterns that suggest, not just attention guidance by target features, but also guidance by scene context (e.g., fixations along walls in the search of clocks). Using machine learning and the psychologically-meaningful principle of reward, it is possible to learn the visual features used in goal-directed attention control.

Schlagwörter

Stichwörter der Autor:innen

attention modelseye behaviorfixation predictionreinforcement learningtop-down attentionvisual search

Autor:innen und Institutionen

  • Department of Psychology, Stony Brook University, Stony Brook, NY, 11794, USA.
  • Department of Computer Science, Stony Brook University, Stony Brook, NY, 11794, USA.

Förderung

  • NEI NIH HHS

Zitieren

APA

Zelinsky, G. J., Chen, Y., Ahn, S., Adeli, H., Yang, Z., Huang, L., Samaras, D., & Hoai, M. (2021). Predicting Goal-directed Attention Control Using Inverse-Reinforcement Learning. Neurons, Behavior, Data analysis, and Theory. https://doi.org/10.51628/001c.22322

Vancouver

Zelinsky GJ, Chen Y, Ahn S, Adeli H, Yang Z, Huang L, et al. Predicting Goal-directed Attention Control Using Inverse-Reinforcement Learning. Neurons, Behavior, Data analysis, and Theory. 2021. doi:10.51628/001c.22322

BibTeX
@article{Zelinsky2021,
  author = {Zelinsky, Gregory J and Chen, Yupei and Ahn, Seoyoung and Adeli, Hossein and Yang, Zhibo and Huang, Lihan and Samaras, Dimitrios and Hoai, Minh},
  title = {Predicting Goal-directed Attention Control Using Inverse-Reinforcement Learning},
  journal = {Neurons, Behavior, Data analysis, and Theory},
  year = {2021},
  doi = {10.51628/001c.22322}
}
Kostenlos starten →