Hybrid deep learning with optimal feature selection for speech emotion recognition using improved meta-heuristic algorithm

Kotha, Manohar; Логашанмугам, Э.

doi:10.1016/j.knosys.2022.108659

Cited by 33 publications

(5 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RNN remembers previous information through self connection and has an impact on the output of subsequent nodes. Therefore, it has better expressive power when processing temporal data than other types of NNs such as fully connected networks and CNNs 13 .…”

Section: Application Of Cnn and Rnn In Srmentioning

confidence: 99%

Application practice of neural network algorithms in speech recognition technology

Guo

2024

Second International Conference on Physics, Photonics, and Optical Engineering (ICPPOE 2023)

View full text Add to dashboard Cite

Speech recognition (SR) technology, as one of the core technologies of human-computer interaction, aims to enable computers to understand the process of converting speech signals into corresponding text or commands through natural language. With the exponential increase of internet information, the features of massive speech data have significant non-specific differences and noise interference. Common feature extraction and transformation methods are no longer sufficient to meet the current needs of model training and recognition. With the rapid growth of machine learning (ML), many researchers use neural networks (NN) to solve various problems in the SR field. This article designs a deep learning (DL) algorithm based on convolutional neural networks (CNN) and recurrent neural networks (RNN) for SR. Firstly, sample filtering, pre weighting, signal framing, and endpoint detection are performed on the speech signal. Secondly, the MFCC value of the preprocessed data is extracted. Finally, an NN model is trained and constructed, and the trained qualified model is used to complete the recognition of speech features. The experimental results show that the algorithm designed in this paper has a lower error rate for SR and stronger generalization ability, which is of great significance for the study of SR.

show abstract

Section: Application Of Cnn and Rnn In Srmentioning

confidence: 99%

Application practice of neural network algorithms in speech recognition technology

Guo

2024

Second International Conference on Physics, Photonics, and Optical Engineering (ICPPOE 2023)

View full text Add to dashboard Cite

show abstract

“…Chen et al [8] used a portion of the IEMOCAP dataset for three different models: a shallow CNN combined with a Bi-LSTM, a deep CNN with a shallow Bi-LSTM, and a deep CNN with a deep Bi-LSTM. Manohar and Logashanmugam [9] integrated different meta-heuristic and deep-learning methods for SER with selected features. Wen et al [10] Introduced self-labeling feature frames in their DLbased SER study.…”

Section: Literature Reviewmentioning

confidence: 99%

Emotion Recognition with Intensity Level from Bangla Speech using Feature Transformation and Cascaded Deep Learning Model

Billah,

Sarker,

Akhand

et al. 2024

IJACSA

View full text Add to dashboard Cite

Speech Emotion Recognition (SER) identifies and categorizes emotional states by analyzing speech signals. The intensity of specific emotional expressions (e.g., anger) conveys critical directives and plays a crucial role in social behavior. SER is intrinsically language-specific; this study investigated a novel cascaded deep learning (DL) model to Bangla SER with intensity level. The proposed method employs the Mel-Frequency Cepstral Coefficient, Short-Time Fourier Transform (STFT), and Chroma STFT signal transformation techniques; the respective transformed features are blended into a 3D form and used as the input of the DL model. The cascaded model performs the task in two stages: classify emotion in Stage 1 and then measure the intensity in Stage 2. DL architecture used in both stages is the same, which consists of a 3D Convolutional Neural Network (CNN), a Time Distribution Flatten (TDF) layer, a Long Short-term Memory (LSTM), and a Bidirectional LSTM (Bi-LSTM). CNN first extracts features from 3D formed input; the features are passed through the TDF layer, Bi-LSTM, and LSTM; finally, the model classifies emotion along with its intensity level. The proposed model has been evaluated rigorously using developed KBES and other datasets. The proposed model revealed as the best-suited SER method compared to existing prominent methods achieving accuracy of 88.30% and 71.67% for RAVDESS and KBES datasets, respectively.

show abstract

“…The recordings contain fifteen phonetically balanced sentences for each emotion: two emotion-specific, three common, and ten cliched phrases. Wavelet-based feature extraction techniques [127], DL strategies [128], gradient boosting classifiers [129], and multisource information fusion have all been tested with it [130][131].…”

Section: Important Databases For Sermentioning

confidence: 99%

A State-of-arts Review of Deep Learning Techniques for Speech Emotion Recognition

Kshirod Sarmah

2024

jes

View full text Add to dashboard Cite

In sophisticated Human-Computer Interfaces (HCI), the emotional state of the user is becoming a crucial component that is closely linked to emotional speech recognition. Spoken expressions, which can be a part of human-machine interaction, are an important source of emotional information. Speech emotion recognition (SER) in deep learning (DL) continues to be a hot topic, especially in the field of emotional computing. Current deep learning (DL) and neural network methods are applied in this highly active field of research. This is as a result of its expanding potential, advancements in algorithms, and practical uses. Quantitative factors such as pitch, intensity, accent and Mel-Frequency Cepstral Coefficients (MFCC) can be employed to model the paralinguistic data contained in human speech. To achieve SER, three key procedures are usually followed: data processing, feature selection/extraction, and classification based on the underlying emotional qualities. The nature of these processes and the peculiarities of human speech lend support to the employment of DL techniques for SER implementation. A variety of DL methods have been used for SER tasks in recent affective computing research works; however, only a small number of them capture the underlying ideas and methodologies that can be used to facilitate the three main steps of SER implementation. With a focus on the three SER implementation processes, we provide a state-of-the-art assessment of research conducted over the last ten years that tackled SER tasks from DL perspectives in this work. Various issues are covered in detail, including the problem of low classification accuracy of Speaker-Independent experiments and the related remedies. The review offers principles for SER evaluation as well, emphasizing indicators that can be experimented with and common baselines.

show abstract

Hybrid deep learning with optimal feature selection for speech emotion recognition using improved meta-heuristic algorithm

Cited by 33 publications

References 34 publications

Application practice of neural network algorithms in speech recognition technology

Application practice of neural network algorithms in speech recognition technology

Emotion Recognition with Intensity Level from Bangla Speech using Feature Transformation and Cascaded Deep Learning Model

A State-of-arts Review of Deep Learning Techniques for Speech Emotion Recognition

Contact Info

Product

Resources

About