Affective Recognition for Multimedia Environments A Review

Abstract

-Detecting emotional responses in multimedia environments is an academically and technologically challenging research issue. In the domain of Affective Computing, from non-interactive and static stimuli (e.g. affective image) to highly interactive and dynamic environments (affective virtual realities), researchers have employed a wide range of affective stimuli to measure and interpret human psychological and physiological emotional behaviours. Various psychophysiological parameters (e.g. Electroencephalography, Galvanic Skin Response, Heart Rate, etc.) have been employed and investigated, in order to detect and quantify human affective states. In this paper, we present a detailed literature review of over 33 affective computing studies, undertaken since 1993. All aspects of these studies (stimuli type, pre-processing, windowing, features, classification technique, etc.) have been reported in detail. We believe that this paper not only summarises the breadth of research over the past 20 years, but also serves to clarify various significant aspects and details of this increasingly valuable and relevant research area.

Affective Recognition in Dynamic and Interactive Virtual Environments

“…In the present study, the two most powerful affective games, in each of the four Affective Clusters 6 introduced in [19], have been identified using the Cosine Similarity Algorithm [20] as implemented in [17]. As a result of this analysis, the eight most powerful affective games (those, which have the highest probability of driving the emotional experience of the participants toward all affective clusters) have been identified.…”

“…As discussed in [19], the majority of studies have employed EEG, Heart Rate and GSR signals to perform affective analysis and recognition. Therefore, in the present study it was decided to record data using these three techniques, for the purposes of supporting the psychophysiological database construction process.…”

“…As discussed in [19], majority of affective recognition studies apply no filtering technique to heart rate and GSR signals recorded from the participants. Therefore in this study, we also decided to use the heart rate and GSR signals without applying any filter.…”

“…Therefore in this study, we also decided to use the heart rate and GSR signals without applying any filter. However, and as discussed in [19], majority of previous studies employed a band-pass filter (majority using 4Hz to 45Hz) in their EEG filtering process. Eyeblink artefacts are typically observed in frequencies lower than 4Hz, as humans rarely blink more than 4 times a second.…”

“…As discussed in [19], majority of artefact removal techniques were performed either by the use of Electrooculography (EOG) signals, or by using computationally expensive EEG artefact removal algorithms, such as those based on Least Mean Square (LMS) and Blind Source Separation (BSS). Due to the absence of EOG signal recording in this study, and the high computational expense of other artefact removal algorithms, it was decided to apply NO EEG artefact removal technique.…”