The health disorders due to stress and depression should not be considered trivial because it has a negative impact on health. Prolonged stress not only triggers mental fatigue but also affects physical health. Therefore, we must be able to identify stress early. In this paper, we proposed the new methods for stress recognition on three classes (neutral, low stress, high stress) from a facial frontal image. Each image divided into three parts, i.e. pairs of eyes, nose, and mouth. Facial features have extracted on each image pixel using DoG, HOG, and DWT. The strength of orthonormality features is considered by the RICA. The GDA distributes the nonlinear covariance. Furthermore, the histogram features of the image parts are applied at a depth-based learning of ConvNet to model the facial stress expression. The proposed method is used FERET databases for training and validation. The k-fold validation method is used as a validation with k=5. Based on the experiments result, the proposed method accuracy showing outperforms compared with other works.

This paper has proposed a system that can be analyzed of speech stress recognition. The proposed method (ensemble SVM and NN) is analyzed comparatively proven to have high accuracy. The ensemble method has been applied to improve machine learning ability in identifying with a small number of datasets. It is caused; due to stress is one of the unconscious emotions. Stress can be recognized by speech however it is not robust. It was caused by small datasets. In this work, we use the sample of SUSAS dataset. The dataset is divided into 10 groups by the combinational method. Each group of data trained using SVM then combined it with trained NN. The experimental results show that with a small dataset, the proposed method outperformed the previous methods.

The facial stress recognition with histogram information is discussed in this paper. The patterns of facial stress are recognized into three stages, these include a registered image, feature extraction, and classification. The registered image process takes three important parts of the face (the pair of eyes, the nose, and the mouth). While the histogram method such as Gabor filter and HOG feature is used as the feature extraction. Furthermore, the proposed method is to combine SVM and Slant Binary Tree algorithm. Each part of face processed on a different pipeline. Based on the experiment result, the proposed system has accuracy about 86.7{\% and it is outperforming the previous methods. The results of the tree structure show that the nose is a part of the face that most indicates stress.

To recognize stress and emotion, most of the existing methods only observe and analyze speech patterns from present-time features. However, an emotion (especially for stress) can change because it was triggered by an event while speaking. To address this issue, we propose a novel method for predicting stress and emotions by analyzing prior emotional states. We named this method the deep time-delay Markov network (DTMN). Structurally, the proposed DTMN contains a hidden Markov model (HMM) and a time-delay neural network (TDNN). We evaluated the effectiveness of the proposed DTMN by comparing it with several state transition methods in predicting an emotional state from time-series (sequences) speech data of the SUSAS dataset. The experimental results show that the proposed DTMN can accurately predict present emotional states by outperforming the baseline systems in terms of the prediction error rate (PER). We then modeled the emotional state transition using a finite Markov chain based on the prediction result. We also conducted an ablation experiment to observe the effect of different HMM values and TDNN parameters on the prediction result and the computational training time of the proposed DTMN.

One technique that is commonly used for mobile robots is an inverted pendulum based model. This research has been implementing a mobile robot technique in an unstable environment. The goal is to design and implementing a discrete digital control system that will provide robot stability. The PID controller algorithm and Ensemble Kalman filter (EnKF) implementation would be an ideal test model of this robot. Both of these algorithms are able to improve the performance of control systems. This robot already tested the performance of the PID control system and the EnKF algorithm. The performance of the PID controller algorithm and EnKF is tested by software. The Control system performance is directly dependent on the EnKF algorithm and input parameters of PID controller. Research uses EnKF algorithm and PID controller as a balancing robot. The covariance filter tuned by manually. Experiments carried out by the method of trial and error by varying the process noise covariance matrix. The system overshoot can be reduced by processing noise covariance matrix. The experiment results showed system optimal on Q_accelerometer: 0001, Q_gyroscope: 0.05 R_measurement: 12:03, P = 1790,005, I = 0.129 and D = 96 881.