Machine Learning in Human Emotion Detection from the Speech

The identification of speech emotions is amongst the most strenuous and fascinating fields of machine learning science. In this article, Chinese emotions are classified as a disruptive atmosphere that classifies several feelings into four major emotional organizations: pleasure, sorrow, resentment, and neutrality. A machine learning in human emotion detection (ML-HED) framework is proposed. The technology suggested removing prosodic and spectrum elements of an audio wave, such as a pulse, power, amplitude, Cepstrum melt frequency correlations, linearly fixed Cepstral, and identification with a template. In all, 87,75% of performers’ statements and 93% of women’s actors were given reliability. The research findings show that the revolutionary technology achieves greater precision by accurately interpreting the feelings, which contrasts with current speech emotion recognition approaches. Besides, the derived characteristics were contrasting with various classification techniques in this study for the comprehensive idea.

Evaluation of active power in dielectric barrier discharge with pulse power supply

The paper describes a practical method for real-time measurement of active power in a dielectric barrier discharge by calculating a Lissajous figures based on the Manley formula. Such a method is very useful for controlling the output ozone concentration in ozone generators taking into account the active power in a DBD reactor. A dielectric barrier discharge (DBD) in atmospheric pressure air was excited by a high-voltage pulse power supply. The active power of the reactor was estimated from the area of the Lissajous figure formed by the V-Q traces, which is directly proportional to the energy consumed per cycle at voltages of 8 and 12 kV, respectively.