• ETRI Journal
• /
• v.28 no.4
• /
• pp.502-505
• /
• 2006

A new class-based histogram equalization method is proposed for robust speech recognition. The proposed method aims at not only compensating the acoustic mismatch between training and test environments, but also at reducing the discrepancy between the phonetic distributions of training and test speech data. The algorithm utilizes multiple class-specific reference and test cumulative distribution functions, classifies the noisy test features into their corresponding classes, and equalizes the features by using their corresponding class-specific reference and test distributions. Experiments on the Aurora 2 database proved the effectiveness of the proposed method by reducing relative errors by 18.74%, 17.52%, and 23.45% over the conventional histogram equalization method and by 59.43%, 66.00%, and 50.50% over mel-cepstral-based features for test sets A, B, and C, respectively.

• Phonetics and Speech Sciences
• /
• v.6 no.2
• /
• pp.3-8
• /
• 2014

This paper deals with how mental fatigue has an effect on human voice. For this a monotonous task to increase the feeling of the fatigue and a set of subjective questionnaire for rating the fatigue were designed. From the experiments the designed task was proven to be monotonous based on the results of the questionnaire responses. To investigate a statistical relationship between speech features extracted from the collected speech data and fatigue, the T test for two-related-samples was used. Statistical analysis shows that speech parameters deeply related to the fatigue are the first formant bandwidth, Jitter, H1-H2, cepstral peak prominence, and harmonics-to-noise ratio. According to the experimental results, it can be seen that voice is changed to be breathy as mental fatigue proceeds.

• MALSORI
• /
• no.60
• /
• pp.145-164
• /
• 2006

This paper proposes class histogram equalization (CHEQ) to compensate noisy acoustic features for robust speech recognition. CHEQ aims to compensate for the acoustic mismatch between training and test speech recognition environments as well as to reduce the limitations of the conventional histogram equalization (HEQ). In contrast to HEQ, CHEQ adopts multiple class-specific distribution functions for training and test environments and equalizes the features by using their class-specific training and test distributions. According to the class-information extraction methods, CHEQ is further classified into two forms such as hard-CHEQ based on vector quantization and soft-CHEQ using the Gaussian mixture model. Experiments on the Aurora 2 database confirmed the effectiveness of CHEQ by producing a relative word error reduction of 61.17% over the baseline met-cepstral features and that of 19.62% over the conventional HEQ.

• International Journal of Computer Science & Network Security
• /
• v.23 no.8
• /
• pp.40-48
• /
• 2023

Speech Emotions recognition has become the active research theme in speech processing and in applications based on human-machine interaction. In this work, our system is a two-stage approach, namely feature extraction and classification engine. Firstly, two sets of feature are investigated which are: the first one is extracting only 13 Mel-frequency Cepstral Coefficient (MFCC) from emotional speech samples and the second one is applying features fusions between the three features: Zero Crossing Rate (ZCR), Teager Energy Operator (TEO), and Harmonic to Noise Rate (HNR) and MFCC features. Secondly, we use two types of classification techniques which are: the Support Vector Machines (SVM) and the k-Nearest Neighbor (k-NN) to show the performance between them. Besides that, we investigate the importance of the recent advances in machine learning including the deep kernel learning. A large set of experiments are conducted on Surrey Audio-Visual Expressed Emotion (SAVEE) dataset for seven emotions. The results of our experiments showed given good accuracy compared with the previous studies.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.2
• /
• pp.133-140
• /
• 2010

We improve the performance of cardiac disorder classification by adding new temporal features extracted from continuous heart sound signals. We add three kinds of novel temporal features to a conventional feature based on mel-frequency cepstral coefficients (MFCC): Heart sound envelope, murmur probabilities, and murmur amplitude variation. In cardiac disorder classification and detection experiments, we evaluate the contribution of the proposed features to classification accuracy and select proper temporal features using the sequential feature selection method. The selected features are shown to improve classification accuracy significantly and consistently for neural network-based pattern classifiers such as multi-layer perceptron (MLP), support vector machine (SVM), and extreme learning machine (ELM).

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.9
• /
• pp.79-87
• /
• 2024

This study aims to improve communication for people with hearing impairments by developing artificial intelligence models that recognize and classify emotions from voice data. To achieve this, we utilized three major AI models: CNN-Transformer, HuBERT-Transformer, and Wav2Vec 2.0, to analyze users' voices in real-time and classify their emotions. To effectively extract features from voice data, we applied transformation techniques such as Mel-Frequency Cepstral Coefficient (MFCC), aiming to accurately capture the complex characteristics and subtle changes in emotions within the voice. Experimental results showed that the HuBERT-Transformer model demonstrated the highest accuracy, proving the effectiveness of combining pre-trained models and complex learning structures in the field of voice-based emotion recognition. This research presents the potential for advancements in emotion recognition technology using voice data and seeks new ways to improve communication and interaction for individuals with hearing impairments, marking its significance.

• Journal of the Korean Society of Industry Convergence
• /
• v.21 no.1
• /
• pp.17-27
• /
• 2018

We propose a new approach to control a biped robot motion based on iterative learning of voice command for the implementation of smart factory. The real-time processing of speech signal is very important for high-speed and precise automatic voice recognition technology. Recently, voice recognition is being used for intelligent robot control, artificial life, wireless communication and IoT application. In order to extract valuable information from the speech signal, make decisions on the process, and obtain results, the data needs to be manipulated and analyzed. Basic method used for extracting the features of the voice signal is to find the Mel frequency cepstral coefficients. Mel-frequency cepstral coefficients are the coefficients that collectively represent the short-term power spectrum of a sound, based on a linear cosine transform of a log power spectrum on a nonlinear mel scale of frequency. The reliability of voice command to control of the biped robot's motion is illustrated by computer simulation and experiment for biped walking robot with 24 joint.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.2
• /
• pp.473-476
• /
• 2005

This paper is proposed a robust various feature parameters in noise. Feature parameter MFCC(Mel Frequency Cepstral Coefficient) used in conventional speech recognition shows good performance. But, parameter transformed feature space that uses PCA(Principal Component Analysis)and ICA(Independent Component Analysis) that is algorithm transformed parameter MFCC's feature space that use in old for more robust performance in noise is compared with the conventional parameter MFCC's performance. The result shows more superior performance than parameter and MFCC that feature parameter transformed by the result ICA is transformed by PCA.

• The Journal of the Acoustical Society of Korea
• /
• v.30 no.3
• /
• pp.149-157
• /
• 2011

In this paper, we propose a new cardiac disorder classification method using an support vector machine (SVM) to combine hidden Markov model (HMM) and murmur existence information. Using cepstral features and the HMM Viterbi algorithm, we segment input heart sound signals into HMM states for each cardiac disorder model and compute log-likelihood (score) for every state in the model. To exploit the temporal position characteristics of murmur signals, we divide the input signals into two subbands and compute murmur probability of every subband of each frame, and obtain the murmur score for each state by using the state segmentation information obtained from the Viterbi algorithm. With an input vector containing the HMM state scores and the murmur scores for all cardiac disorder models, SVM finally decides the cardiac disorder category. In cardiac disorder classification experimental results, the proposed method shows the relatively improvement rate of 20.4 % compared to the HMM-based classifier with the conventional cepstral features.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.3
• /
• pp.143-149
• /
• 2020

This paper presents the effect of the feature extraction methods used in the audio preprocessing on the classification performance of the Convolutional Neural Networks (CNN). We extract mel spectrogram, log mel spectrogram, Mel Frequency Cepstral Coefficient (MFCC), and delta MFCC from the UrbanSound8K dataset, which is widely used in environmental sound classification studies. Then we scale the data to 3 distributions. Using the data, we test four CNNs, VGG16, and MobileNetV2 networks for performance assessment according to the audio features and scaling. The highest recognition rate is achieved when using the unscaled log mel spectrum as the audio features. Although this result is not appropriate for all audio recognition problems but is useful for classifying the environmental sounds included in the Urbansound8K.