Gal A. Kaminka: Publications

Sorted by DateClassified by Publication TypeClassified by TopicGrouped by Student (current)Grouped by Former Students

Goal Recognition

Rivka Vizen. Goal Recognition. Master's Thesis, Hebrew University of Jerusalem,2023.

Download

[PDF]1.0MB  

Abstract

The human capacity to recognize another agent's plans and goals allows it to infer missing and future actions and predict the agent's goal. This ability constitutes a fundamental aspect of human social cognition. An in-depth understanding of these cognitive processes serves to identify impairments and promote advanced research in psychology. Additionally, it is relevant to engineering and artificial intelligence, as it can contribute to the development of algorithms for the recognition of goals and plans by computers. Contemporary algorithms for detecting agent intentions operate on the principle of rationality, assuming a predefined list of potential targets. To improve these algorithms to the level of human capacity that has dynamically formulated lists of goals, we need to understand the cognitive and visual attributes utilized by individuals in the goal-selection process. In this study, we propose that the Gestalt Laws serve as a foundational framework for comprehending visual perception's role in goal selection. To test the aforementioned hypothesis, a visual experiment was conducted. Participants were tasked with determining the appropriate placement of an object in space while being surrounded by other objects characterized by Gestalt Law features. We present the experiment's findings, revealing that Gestalt Law does indeed underlie the visual cognitive processes involved in goal selection. In addition, the data collected yields a wealth of information that can be leveraged for further research and exploration of unaddressed questions, such as potential priority distinctions between characteristics.

Additional Information

BibTeX

@mastersThesis{rivki-msc,
  author = {Rivka Vizen},
  title = {Goal Recognition},
  school = {Hebrew University of Jerusalem},
  year = {2023},
  wwwnote = {}, 
  abstract = {The human capacity to recognize another agent's plans and goals allows it to infer missing 
and future actions and predict the agent's goal. This ability constitutes a fundamental 
aspect of human social cognition. An in-depth understanding of these cognitive processes 
serves to identify impairments and promote advanced research in psychology. 
Additionally, it is relevant to engineering and artificial intelligence, as it can contribute to 
the development of algorithms for the recognition of goals and plans by computers. 
Contemporary algorithms for detecting agent intentions operate on the principle of 
rationality, assuming a predefined list of potential targets. To improve these algorithms to 
the level of human capacity that has dynamically formulated lists of goals, we need to 
understand the cognitive and visual attributes utilized by individuals in the goal-selection 
process. In this study, we propose that the Gestalt Laws serve as a foundational framework 
for comprehending visual perception's role in goal selection. 
To test the aforementioned hypothesis, a visual experiment was conducted. Participants 
were tasked with determining the appropriate placement of an object in space while being 
surrounded by other objects characterized by Gestalt Law features. 
We present the experiment's findings, revealing that Gestalt Law does indeed underlie the 
visual cognitive processes involved in goal selection. In addition, the data collected yields 
a wealth of information that can be leveraged for further research and exploration of 
unaddressed questions, such as potential priority distinctions between characteristics. 
  },
}

Generated by bib2html.pl (written by Patrick Riley ) on Fri Aug 30, 2024 17:29:51