A user-centered approach for detecting emotions with low-cost sensors
- PDF / 3,641,542 Bytes
- 23 Pages / 439.642 x 666.49 pts Page_size
- 52 Downloads / 159 Views
A user-centered approach for detecting emotions with low-cost sensors Rita Francese1
· Michele Risi1 · Genoveffa Tortora1
Received: 10 December 2019 / Revised: 20 July 2020 / Accepted: 11 August 2020 / © The Author(s) 2020
Abstract Detecting emotions is very useful in many fields, from health-care to human-computer interaction. In this paper, we propose an iterative user-centered methodology for supporting the development of an emotion detection system based on low-cost sensors. Artificial Intelligence techniques have been adopted for emotion classification. Different kind of Machine Learning classifiers have been experimentally trained on the users’ biometrics data, such as hearth rate, movement and audio. The system has been developed in two iterations and, at the end of each of them, the performance of classifiers (MLP, CNN, LSTM, BidirectionalLSTM and Decision Tree) has been compared. After the experiment, the SAM questionnaire is proposed to evaluate the user’s affective state when using the system. In the first experiment we gathered data from 47 participants, in the second one an improved version of the system has been trained and validated by 107 people. The emotional analysis conducted at the end of each iteration suggests that reducing the device invasiveness may affect the user perceptions and also improve the classification performance. Keywords Emotion detection · Affective analysis · Artificial intelligence · Low-cost sensors
1 Introduction Recently, many research efforts have been devoted to the recognition of human emotions. This interest is mainly due to the fact that emotions impact on the uses’ reactions and behaviours, and their understanding may be useful in various fields, such as humancomputer interaction, software engineering [11], gaming, marketing and multimedia. In Rita Francese
[email protected] Michele Risi [email protected] Genoveffa Tortora [email protected] 1
Department of Computer Science, University of Salerno, Fisciano, Italy
Multimedia Tools and Applications
health-care it is particularly useful for specific types of patients, like autistic people or people which are not able to express their sentiments: to recognize their emotions may be useful for caregivers to prevent panic attacks or to better understand their needs. Emotions are related to the mood of a person and last for few instants, so they are difficult to detect. Many emotion recognition approaches are based on the analysis of the expressions of the user’s face, voice, and gestures, even if these user behaviors may be intentionally controlled or hide other emotions. In addition, the user expressions may also be affected by ethnicity and cultures. These problems may be overcome by adopting emotion recognition approaches based on physiological signals, which are not visible at the human eye and immediately reflect the emotional changes. These kinds of signals may be detected by sensors. The use of low-cost sensors is a relevant additional challenge to let this technology be accessible to all [13]. In addition, it m
Data Loading...
