Thesis
Improving non-player character interaction using physiological data
Master of Science (M.S.), Drexel University
Jun 2016
DOI:
https://doi.org/10.17918/etd-6863
Abstract
Non-player characters (NPCs) in video games have very little information about the player's current state. This disconnect leads to less than ideal interactions between the human and the computer. Measurements of a human's physiological state have been used to drive a wide range of software interactions, such as biofeedback applications. The usage of physiological data in games has been very limited, mainly to adjustments in difficulty based on stress levels. However, research with virtual agents has shown how useful physiological data can be in human-computer interaction. Based on these findings, this thesis assesses the usefulness of physiological signals for the interaction with NPCs. Measurements of skin conductance on the fingers and facial muscle tension serves as a means to estimate the player's emotional state at any given time. This data is then used to adjust the behavior of non-player characters in so far as their dialogue acknowledges the player's emotion. An experimental evaluation of the developed system showed that using a combination of electromyography and electrodermal activity to estimate human emotion affords non-player characters with more information about the player, demonstrating the viability of the approach. In the small sample of the evaulation, there was no significant difference for a questionnaire measure of rapport with the NPCs. However, the qualitative feedback in the questionnaires showed a clear difference in the perception of the system's use of physiological information. How this information can be used most effectively by non-player characters should be explored further in future research.
Metrics
44 File views/ downloads
19 Record Views
Details
- Title
- Improving non-player character interaction using physiological data
- Creators
- Joe Jalbert - DU
- Contributors
- Stefan Rank (Advisor) - Drexel University (1970-)Jichen Zhu (Advisor) - Drexel University (1970-)Santiago Ontañón (Advisor) - Drexel University (1970-)
- Awarding Institution
- Drexel University
- Degree Awarded
- Master of Science (M.S.)
- Publisher
- Drexel University; Philadelphia, Pennsylvania
- Number of pages
- viii, 44 pages
- Resource Type
- Thesis
- Language
- English
- Academic Unit
- Digital Media; Drexel University; Antoinette Westphal College of Media Arts and Design
- Other Identifier
- 6863; 991014632563404721