Support Vector Slant Binary Tree Architecture for Facial Stress Recognition Based on Gabor and HOG Feature

Prasetio, Barlian Henryranu; Tamura, H.; Tanno, Koichi

doi:10.1109/iwbis.2018.8471709

Cited by 9 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In one popular method of recognizing stress using facial images, unlike the facial action unit, a comprehensive feature is extracted from the entire image. In some studies, the HOG features were extracted from the eye, nose, and mouth regions in RGB images and used as features [26,29]. In these methods, a CNN and a method combining the SVM and slant binary tree algorithm were used as classifiers.…”

Section: Facial-image-based Stress Recognition Methodsmentioning

confidence: 99%

“…By contrast, images, such as thermal images showing blood flow and respiratory rate and visual images portraying body movements and pupil size, can be used for stress recognition [22][23][24]. Some stress-recognition studies use only visual images, especially facial images, which have the advantage of only requiring a camera; the subjects need not wear additional equipment [25,26]. However, in many of these methods, handcrafted features continue to be used.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep-Learning-Based Stress Recognition with Spatial-Temporal Facial Information

Jeon

Bae

Lee

et al. 2021

Sensors

View full text Add to dashboard Cite

In recent times, as interest in stress control has increased, many studies on stress recognition have been conducted. Several studies have been based on physiological signals, but the disadvantage of this strategy is that it requires physiological-signal-acquisition devices. Another strategy employs facial-image-based stress-recognition methods, which do not require devices, but predominantly use handcrafted features. However, such features have low discriminating power. We propose a deep-learning-based stress-recognition method using facial images to address these challenges. Given that deep-learning methods require extensive data, we constructed a large-capacity image database for stress recognition. Furthermore, we used temporal attention, which assigns a high weight to frames that are highly related to stress, as well as spatial attention, which assigns a high weight to regions that are highly related to stress. By adding a network that inputs the facial landmark information closely related to stress, we supplemented the network that receives only facial images as the input. Experimental results on our newly constructed database indicated that the proposed method outperforms contemporary deep-learning-based recognition methods.

show abstract

Section: Facial-image-based Stress Recognition Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep-Learning-Based Stress Recognition with Spatial-Temporal Facial Information

Jeon

Bae

Lee

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…The literature regarding stress detection is extensive, and very often constitutes a part of emotion recognition process, since stress can be reflected through subject's emotional state or mood such as frustration, anger, agitation, preoccupation, fear, anxiety, and tenseness [63]. As in case of emotion recognition, it is mainly based on facial [63], [64], speech [65], [66] and gestures [67]- [69] signals analysis. The undoubted benefit of this approach is the possibility of interpreting the through less or nonintrusive methods which do not require physical contact [63].…”

Section: B Stress Recognition Using Physiological Datamentioning

confidence: 99%

Stress Reduction Using Bilateral Stimulation in Virtual Reality

et al. 2020

View full text Add to dashboard Cite

The goal of this research is to integrate Virtual Reality (VR) with the bilateral stimulation used in EMDR as a tool to relieve stress. We created a 15 minutes relaxation training program for adults in a virtual, relaxing environment in form of a walk in the woods. The target platform for the tool is HTC Vive, however it can be easily ported to other VR platforms. An integral part of this tool is a set of sensors, which serves as physiological measures to evaluate the effectiveness of such system. What is more, the system integrate visual (passing sphere), auditory (surround sound) and tactile signals (vibration of controllers). A pilot treatment programme, incorporating the above mentioned VR system, was carried out. Experimental group consisting of 28 healthy adult volunteers (office workers), participated in three different sessions of relaxation training. Before starting, baseline features such as subjectively perceived stress, mood, heart rate, galvanic skin response and muscle response were registered. The monitoring of physiological indicators is continued during the training session and one minute after its completion. Before and after the session, volunteers were asked to re-fill questionnaires regarding the current stress level and mood. The obtained results were analyzed in terms of variability over time: before, during and after the session.

show abstract

“…Observing that the signs of stress could be more easily detected by looking at the condition of the face, particularly the lines or wrinkles around the nose, mouth, and eyes, [ 22 , 23 ] investigated three facial parts (the eyes, nose and mouth) which are significant for stress detection. [ 23 ] extracted Gabor filter and HOG (Histogram of Oriented Gradients) features from each part of the face in pixels through visual image encoding process, and fed them into three different SVM classifiers. The obtained three results were then fed into slant binary tree to get the final results.…”

Section: Related Workmentioning

confidence: 99%

Video-Based Stress Detection through Deep Learning

Zhang

Feng

et al. 2020

Sensors

View full text Add to dashboard Cite

Stress has become an increasingly serious problem in the current society, threatening mankind’s well-beings. With the ubiquitous deployment of video cameras in surroundings, detecting stress based on the contact-free camera sensors becomes a cost-effective and mass-reaching way without interference of artificial traits and factors. In this study, we leverage users’ facial expressions and action motions in the video and present a two-leveled stress detection network (TSDNet). TSDNet firstly learns face- and action-level representations separately, and then fuses the results through a stream weighted integrator with local and global attention for stress identification. To evaluate the performance of TSDNet, we constructed a video dataset containing 2092 labeled video clips, and the experimental results on the built dataset show that: (1) TSDNet outperformed the hand-crafted feature engineering approaches with detection accuracy 85.42% and F1-Score 85.28%, demonstrating the feasibility and effectiveness of using deep learning to analyze one’s face and action motions; and (2) considering both facial expressions and action motions could improve detection accuracy and F1-Score of that considering only face or action method by over 7%.

show abstract

Support Vector Slant Binary Tree Architecture for Facial Stress Recognition Based on Gabor and HOG Feature

Cited by 9 publications

References 10 publications

Deep-Learning-Based Stress Recognition with Spatial-Temporal Facial Information

Deep-Learning-Based Stress Recognition with Spatial-Temporal Facial Information

Stress Reduction Using Bilateral Stimulation in Virtual Reality

Video-Based Stress Detection through Deep Learning

Contact Info

Product

Resources

About