Ensemble Learning by High-Dimensional Acoustic Features for Emotion Recognition from Speech Audio Signal

Chalapathi, M. M. Venkata; Kumar, Manoj; Sharma, Neeraj; Shitharth, S.

doi:10.1155/2022/8777026

Cited by 18 publications

(5 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…EV is the metaclassifier that determines the final prediction result through voting [34]. It is the last layer before the preclassification layer, which is composed of the LRC, RFC, GBC, and multilayer perceptron classifier (MLP-C).…”

Section: Ensemble Votingmentioning

confidence: 99%

Feature Aggregation with Two‐Layer Ensemble Framework for Multilingual Speech Emotion Recognition

Ough,

Pyo,

Kim

2023

Mathematical Problems in Engineering

View full text Add to dashboard Cite

In this study, we present a framework for improving the accuracy of speech emotion recognition in a multilingual environment. In our prior experiments, where machine learning (ML) models were trained to predict emotions in Korean and then tested in English, as well as vice versa, we observed a dependency on language in emotion recognition, resulting in poor accuracy. We suspect that this may be related to the spectral differences in certain emotions between Korean and English and to the tendency for different formant values to have different acoustic frequencies. For this study, we investigated several different methods, including models with mixed databases, a single database, and bagging, boosting, and voting ML algorithms. Finally, we developed a framework consisting of two branches: one for the aggregation of high-dimensional features from multilingual data and one for a two-layered ensemble framework for emotion classification. In the ensemble framework for Korean and English (EF-KEN), features are extracted and ensemble models are trained, boosted, and evaluated by applying them to different spoken languages (English and Korean). The final experimental result demonstrates a meaningful improvement in an environment with two different languages.

show abstract

Section: Ensemble Votingmentioning

confidence: 99%

Feature Aggregation with Two‐Layer Ensemble Framework for Multilingual Speech Emotion Recognition

Ough,

Pyo,

Kim

2023

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…The function, which uses binary input to evaluate trustworthiness, has recently become one of conventional strategies. It counts the true and false sensing of the user has engaged in before using probability functions to calculate the trusted sensing value [42], [43]. [44] Reviewing the related works, the complexities pertaining to accuracy in the crowdsourcing schemes are evident, and despite that, some of the existing frameworks or solutions are addressing certain issues.…”

Section: Trust Mechanismmentioning

confidence: 99%

“…root-mean-square-error[42] of the Degree-of-fair-cooperation () DFC csp of the crowd sensor pool csp assesses as follows in (Eq 10): csp denotes the count of crowd-sensors found in crowd sensor pool csp , {} csp denotes the list of crowd-sensors used to establish the crowd sensor pool cspIn furtherance to the above process, the optimal crowd sensor pool among the set of crowd sensor pools CSP can be assessed as:• The crowd sensor pools in CSP shall be sorted in the order of descending manner as per the() DFC csp value • Set of crowd sensor pools that are with greater () DFC csp value than the threshold value  shall be sorted • Also, the crowd sensor pools discovered in crowd sensor pool request process that is having Degreeof-fair-cooperation () DFC csp more than the threshold value  shall also be sorted International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169 Volume: 10 Issue: 10 DOI: https://doi.org/10.17762/ijritcc.v10i10.5737 Article Received: 28 July 2022 Revised: 05 September 2022 Accepted: 12 September 2022 ___________________________________________________________________________________________________________________ 71 IJRITCC | October 2022, Available @ http://www.ijritcc.org…”

mentioning

confidence: 99%

Affirmed Crowd Sensor Selection based Cooperative Spectrum Sensing

Rao¹,

Prasad²,

Prasad³

2022

IJRITCC

View full text Add to dashboard Cite

The Cooperative Spectrum sensing model is gaining importance among the cognitive radio network sharing groups. While the crowd-sensing model (technically the cooperative spectrum sensing) model has positive developments, one of the critical challenges plaguing the model is the false or manipulated crowd sensor data, which results in implications for the secondary user’s network. Considering the efficacy of the spectrum sensing by crowd-sensing model, it is vital to address the issues of falsifications and manipulations, by focusing on the conditions of more accurate determination models. Concerning this, a method of avoiding falsified crowd sensors from the process of crowd sensors centric cooperative spectrum sensing has portrayed in this article. The proposal is a protocol that selects affirmed crowd sensor under diversified factors of the decision credibility about spectrum availability. An experimental study is a simulation approach that evincing the competency of the proposal compared to the other contemporary models available in recent literature.

show abstract

“…The overall performance of the ensemble method using a majority vote was lower as compared to individual classifiers. Chalapathi et al [26] proposed ensemble learning using high-dimensional acoustic features for audio emotion recognition. The AdaBoost classifier was used for classification.…”

Section: Introductionmentioning

confidence: 99%

Audio Based Emotion Classification Using Classifier Ensemble

Mudassar,

Ul Haq,

Majid

et al. 2023

PJETS

View full text Add to dashboard Cite

This paper presents a novel approach of combining classifiers outputs for audio emotion recognition. The proposed classifiers ensemble technique combines the confusion matrices of base classifiers. It is because some classifiers with overall lower performance have better accuracy for a specific class as compared to others with overall higher accuracy. In this approach, the best results obtained for different emotion classes from various classifiers are combined to create a combined confusion matrix. The performance of this approach was analyzed using three emotional speech databases in different languages, i.e., Berlin emotional speech database (EMO-DB), Italian emotional speech database (EMOVO-DB), and Surrey audio-visual expressed emotion database (SAVEE-DB). The openSMILE toolkit was used to extract a total of 8543 audio features. These features include pitch, energy, intensity, jitter, shimmer, formants, MFCC, MFB, LSP and spectral features. These features were normalized using min-max normalization technique, while correlation-based feature selection (CFS) with best-first search approach was used for feature reduction. The classification was performed using five different base classifiers, i.e., SVM, MLP, IBK, AdaBoost, and Random Forest. The experimental results showed better performance for the proposed technique as compared to other state-of-the-art methods. The classification accuracies obtained for seven emotion classes were 91.8%, 83.7%, and 80.5% for the EMO-DB, EMOVO-DB, and SAVEE-DB, respectively.

show abstract

Ensemble Learning by High-Dimensional Acoustic Features for Emotion Recognition from Speech Audio Signal

Cited by 18 publications

References 36 publications

Feature Aggregation with Two‐Layer Ensemble Framework for Multilingual Speech Emotion Recognition

Feature Aggregation with Two‐Layer Ensemble Framework for Multilingual Speech Emotion Recognition

Affirmed Crowd Sensor Selection based Cooperative Spectrum Sensing

Audio Based Emotion Classification Using Classifier Ensemble

Contact Info

Product

Resources

About