Description
These data were collected as part of a PhD project 'Using affective computing to make virtual agents more believable and enhance the social interaction with users'.
This study presents the ANX dataset, a mutimodal affective dataset containing anxiety-related physiological data, gameplay videos, face recordings, and self-reported annotations collected during 25-minute gameplay sessions of the commercial horror game "The Evil Within" (Bethesda, 2014). Physiological data relates to Electrodermal activity (ECG), Blood Volume Pulse (BVP), Respiration (RSP), and Galvanic Skin Response (GSR). Annotation data consists of a continuous and unbounded trace of the Arousal affective dimension and discrete Self-assessment mannikins rates of the three affective dimensions Pleasure, Arousal and Dominance, collected using the annotation tool AffectRankTrace. Details on the data can be found in "EMS_4666\data_structure.txt".
The ANX dataset includes data from 26 participants and face recordings from 13 participants, totalling 10 hours and 50 minutes of recordings. The extracted face features are shared as open access data. Access to the face recordings are restricted but may be shared with other researchers on request.
The gameplay content is the intellectual property of the video game publisher Bethesda Softworks, and is subject to their End-User License Agreement (EULA). Due to copyright and licensing restrictions, raw gameplay footage is also available on request only. Please use the contact details to request access to the face recordings and the gameplay videos. All requests will be subject to an approval process.
This study presents the ANX dataset, a mutimodal affective dataset containing anxiety-related physiological data, gameplay videos, face recordings, and self-reported annotations collected during 25-minute gameplay sessions of the commercial horror game "The Evil Within" (Bethesda, 2014). Physiological data relates to Electrodermal activity (ECG), Blood Volume Pulse (BVP), Respiration (RSP), and Galvanic Skin Response (GSR). Annotation data consists of a continuous and unbounded trace of the Arousal affective dimension and discrete Self-assessment mannikins rates of the three affective dimensions Pleasure, Arousal and Dominance, collected using the annotation tool AffectRankTrace. Details on the data can be found in "EMS_4666\data_structure.txt".
The ANX dataset includes data from 26 participants and face recordings from 13 participants, totalling 10 hours and 50 minutes of recordings. The extracted face features are shared as open access data. Access to the face recordings are restricted but may be shared with other researchers on request.
The gameplay content is the intellectual property of the video game publisher Bethesda Softworks, and is subject to their End-User License Agreement (EULA). Due to copyright and licensing restrictions, raw gameplay footage is also available on request only. Please use the contact details to request access to the face recordings and the gameplay videos. All requests will be subject to an approval process.
| Date made available | 8 Oct 2025 |
|---|---|
| Publisher | Abertay University |
| Date of data production | Jun 2022 - Nov 2022 |
Student theses
-
Using affective computing to enhance the believability of virtual agents and the social interaction with users
Graja, S. (Author), Lovell, G. (Supervisor) & Fee, K. (Supervisor), 25 Feb 2025Student thesis: Doctoral Thesis