Emotional healthcare system: Emotion detection by facial expressions using Japanese database

Geometric positions and optical flow based emotion detection using MLP and reduced dimensions

“…Our system is observed to perform better considering the reduced dimensions and hybrid methodology when evaluated on JAFFE dataset. However, the work by Tivatansakul et al [47] is seen to have performed better than the distance‐wrinkle based methodologies alone.…”

Employing Machine Learning for Multi-perspective Emotional Health Analysis

“…This geometry comprises of shapes, points, and locations of facial components like eyes, nose, eyebrows, and mouth and their distances. The downside of this geometric based approach is that it requires an accurate and reliable detection method (Tivatansakul & Ohkura, 2014 ). Another challenge for geometric-based approach is that the facial geometry of human is affected by the environment and climatic conditions.…”

“…Each region is then encoded into curves, edges and other local features using comparison with neighboring pixels and center value. Thereafter, it constructs a 256 level histogram for each region which are later concatenated to form a global description of entire face (Tivatansakul & Ohkura, 2014 ). Similarly, there exists a Local Directional Pattern (LDP) method which employs edge detector to evaluate edge response in all eight directions.…”

“…Here, it is worth mentioning that the emotional health of ordinary people is also equally concerning and thus must be focused upon (Tivatansakul & Ohkura, 2014 ). Now, it must be realized that society is in dire need of developing systems that aid people to get rid of mental stress and negative emotions (Ambarkar & Akhare, 2020 ).Various researchers have devised different strategies to improve emotional health of patients by providing effective, intelligent, and attractive healthcare systems.…”

“…All these proposed systems understand that the most crucial part of emotional health management is analysis and recognition of real-time emotions which is capable of understanding user emotions in minimum time, so that appropriate corrective therapies could be activated or suggested. Recognition of emotional health is a challenging part as it involves recognition of facial expressions, speech, text, or biological signals for enhanced accuracy (Tivatansakul & Ohkura, 2014 ). According to a popular psychological theory, emotion may be defined as a complex psychological state that involves three distinct components: a subjective experience, a physiological response, and a behavioral or expressive response.…”