Emotion recognition in speech signals using optimization based multi-SVNN classifier

Mannepalli, Kasiprasad; Sastry, Panyam Narahari; Maloji, Suman

doi:10.1016/j.jksuci.2018.11.012

Cited by 30 publications

(23 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MFCC here simplices the human hearing merits; and the linear spectrum is first mapped to the MFCC's nonlinear spectrum based on auditory perception, and then being converted to the Cepstrum. The formula for converting ordinary frequency to Mel frequency is, ( ) 2595*lg(1 / 700) mel f f =+ (12) Then, we pass the spectrum through a set of Mel filters to get the Mel spectrum. The formula is,…”

Section: Features Extraction For Acoustic Signalsmentioning

confidence: 99%

“…The fundamental frequency contains a large number of features that characterize speech affective, which are crucial in acoustic affective recognition. The change range is 50-500Hz; and the detection difficulty is relatively high [12,13]. Commonly, fundamental frequency feature extraction methods are used for the autocorrelation function (ACF)-time domain, average amplitude difference method (AMFD)time domain and wavelet method (WM)-frequency domain; C.K, Y., et al selected higher order spectral features in a set for affective recognition by using 28 bi-spectral features and 22 bi-coherence features [14].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Simplified inverse filter tracked affective acoustic signals classification incorporating deep convolutional neural networks

Kuang

Wang

et al. 2020

Applied Soft Computing

View full text Add to dashboard Cite

Facial expressions, verbal, behavioral, such as limb movements, and physiological features are vital ways for affective human interactions. Researchers have given machines the ability to recognize affective communication through the above modalities in the past decades. In addition to facial expressions, changes in the level of sound, strength, weakness, and turbulence will also convey affective. Extracting affective feature parameters from the acoustic signals have been widely applied in customer service, education, and the medical field. In this research, an improved AlexNet-based deep convolutional neural network (A-DCNN) is presented for acoustic signal recognition. Firstly, preprocessed on signals using simplified inverse filter tracking (SIFT) and short-time Fourier transform (STFT), Mel frequency Cepstrum (MFCC) and waveformbased segmentation were deployed to create the input for the deep neural network (DNN), which was applied widely in signals preprocess for most neural networks. Secondly, acoustic signals were acquired from the public Ryerson Audio-Visual Database of Affective Speech and Song (RAVDESS) affective speech audio system. Through the acoustic signal preprocessing tools, the basic features of the kind of sound signals were calculated and extracted. The proposed DNN based on improved AlexNet has a 95.88% accuracy on classifying eight affective of acoustic signals. By comparing with some linear classifications, such as decision table (DT) and Bayesian inference (BI) and other deep neural networks, such as AlexNet+SVM, recurrent convolutional neural network (R-CNN), etc., the proposed method achieves high effectiveness on the accuracy (A), sensitivity (S1), positive predictive (PP), and f1-score (F1). Acoustic signals affective recognition and classification can be potentially applied in industrial product design through measuring consumers' affective responses to products; by collecting relevant affective sound data to understand the popularity of the product, and furthermore, to improve the product design and increase the market responsiveness.

show abstract

Section: Features Extraction For Acoustic Signalsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Simplified inverse filter tracked affective acoustic signals classification incorporating deep convolutional neural networks

Kuang

Wang

et al. 2020

Applied Soft Computing

View full text Add to dashboard Cite

show abstract

“…There are some other emotion recognition algorithms which use multi SVNN classifiers (Multiple Support Vector Neural Network) [19] for identifying the emotions and sentiment analysis [20]. The significance of obtaining student intellectual states information in an unobtrusive way had incited the headway of various methodology to propel different procedures to advance the result in estimating head posture and analysis of facial expressions.…”

Section: Related Workmentioning

confidence: 99%

Deep Learning-Based Emotion-Sensitive Learning Cognitive State Analysis

2020

IJETER

View full text Add to dashboard Cite

The cognitive state of a student is unique and exciting. The current developments in deep learning gives an unparalleled chance to analysts to assess the cognitive state. But most of the existing cognitive state strategies centre around attention, overlooking the importance of emotions in humans. Human feelings play a huge role in the computer vision industry and numerous researches are performed with its assistance. Hence, our aim is to propose a cognitive state investigation system which is emotion sensitive. It will consequently assess the students' attention based on head posture and emotion recognized from the face detection in an unobtrusive way. The system presented is an implementation of multiple tasks learning cascaded with a convolutional neural network (CNN), introduced for detecting expression, locating landmarks, and estimating head pose all at a time. The expression detection and landmark location help in alignment of the face. The estimation of head pose and face alignment are further used to evaluate the leaner's attention. Exploratory outcomes show that this technique can obtain students' emotion with an accuracy of 94%.

show abstract

“…Many researchers are using a combination of various features of speech signal. Some of them are temporal features [10][11][12].MFCC features [13], pitch Chroma, spectral flux and tonal power ratio [14][15][16]. To classify the signals for a particular application a classification technique will be used [17][18][19][20][21].…”

Section: Introductionmentioning

confidence: 99%

Copy and Move Detection in Audio Recordings using Dynamic Time Warping Algorithm

Mannepalli*¹,

Krishna²,

Krishna³

et al. 2019

IJITEE

View full text Add to dashboard Cite

Copy and move forgery technique is a versatile technique used by criminals to change the evidences unlawfully. This is done by removing or adding the segments of an audio recording from another audio using simple software tools . Nowa-days in courts and forensics lab we use digital audio or speech as proof of evidence. With the advancement in digital software and technology it is made possible to modify the original audio data and tamper with it. In this paper we are using a robust method of copy and move detection using MFCC and Dynamic time warping which is robust against several kinds of regularly used post-processing actions and background noise, which highlights its promising potential of the suggested method as an audio tampering detection tool in various practical forensics applications with help of DTW. The proposed method is giving good outcome in differentiating the original sequences and forged sequences. This is done and implemented in mat lab

show abstract

Emotion recognition in speech signals using optimization based multi-SVNN classifier

Cited by 30 publications

References 22 publications

Simplified inverse filter tracked affective acoustic signals classification incorporating deep convolutional neural networks

Simplified inverse filter tracked affective acoustic signals classification incorporating deep convolutional neural networks

Deep Learning-Based Emotion-Sensitive Learning Cognitive State Analysis

Copy and Move Detection in Audio Recordings using Dynamic Time Warping Algorithm

Contact Info

Product

Resources

About