Terahertz imaging and time-domain spectroscopy have been widely used to characterize the properties of test samples in various biomedical and engineering fields. Many of these tasks require the analysis of acquired terahertz signals to extract embedded information, which can be achieved using machine learning. Recently, machine learning techniques have developed rapidly, and many new learning models and learning algorithms have been investigated. Therefore, combined with state-of-the-art machine learning techniques, terahertz applications can be performed with high performance that cannot be achieved using modeling techniques that precede the machine learning era. In this review, we introduce the concept of machine learning and basic machine learning techniques and examine the methods for performance evaluation. We then summarize representative examples of terahertz imaging and time-domain spectroscopy that are conducted using machine learning.

https://www.mdpi.com/1424-8220/21/4/1186/pdf

Machine Learning Techniques for THz Imaging and Time-Domain Spectroscopy.

We demonstrate an automatic recognition strategy for terahertz (THz) pulsed signals of breast invasive ductal carcinoma (IDC) based on a wavelet entropy feature extraction and a machine learning classifier. The wavelet packet transform was implemented into the complexity analysis of the transmission THz signal from a breast tissue sample. A novel index of energy to Shannon entropy ratio (ESER) was proposed to distinguish different tissues. Furthermore, the principal component analysis (PCA) method and machine learning classifier were further adopted and optimized for automatic classification of the THz signal from breast IDC sample. The areas under the receiver operating characteristic curves are all larger than 0.89 for the three adopted classifiers. The best breast IDC recognition performance is with the precision, sensitivity and specificity of 92.85%, 89.66% and 96.67%, respectively. The results demonstrate the effectiveness of the ESER index together with the machine learning classifier for automatically identifying different breast tissues.

Automatic recognition of breast invasive ductal carcinoma based on terahertz spectroscopy with wavelet packet transform and machine learning

In this paper, we introduced a voice control intelligent wheelchair movement using Convolutional Neural Networks (CNNs). The intelligent wheelchair used four voice commands such as stop, go, left and right to assist disable people to move. Data are collected from google in the wav format. Mel-Frequency Cepstral Coefficient (MFCC) is applied to extract the command voice. The hardware used to deploy the system is Raspberry PI 3B+. The proposed method is using CNNs to classify the voice command and achieved excellent result with 95.30% accuracy. Therefore, the method can be commercialized and hopefully can give benefit to the disable society.

Voice Control Intelligent Wheelchair Movement Using CNNs

In this paper, we develop an intelligent wheelchair using CNNs and SVM voice recognition methods. The data is collected from Google and some of them are self-recorded. There are four types of data to be recognized which are go, left, right, and stop. Voice data are extracted using MFCC feature extraction technique. CNNs and SVM are then used to classify and recognize the voice data. The motor driver is embedded in Raspberry PI 3B+  to control the movement of the wheelchair prototype. CNNs produced higher accuracy i.e. 95.30% compared to SVM which is only 72.39%. On the other hand, SVM only took 8.21 seconds while CNNs took 250.03 seconds to execute. Therefore, CNNs produce better result because noise are filtered in the feature extraction layer before classified in the classification layer. However, CNNs took longer time due to the complexity of the networks and the less complexity implementation in SVM give shorter processing time.

http://ijai.iaescore.com/index.php/IJAI/article/download/20492/pdf

Comparison of CNNs and SVM for voice control wheelchair

Audio signal processing is the most challenging field in the current era for an analysis of an audio signal. Audio signal classification (ASC) comprises of generating appropriate features from a sound and utilizing these features to distinguish the class the sound is most likely to fit. Based on the application’s classification domain, the characteristics extraction and classification/clustering algorithms used may be quite diverse. The paper provides the survey of the state-of art for understanding ASC’s general research scope, including different types of audio; representation of audio like acoustic, spectrogram; audio feature extraction techniques like physical, perceptual, static, dynamic; audio pattern matching approaches like pattern matching, acoustic phonetic, artificial intelligence; classification, and clustering techniques. The aim of this state-of-art paper is to produce a summary and guidelines for using the broadly used methods, to identify the challenges as well as future research directions of acoustic signal processing.

Pattern analysis based acoustic signal processing: a survey of the state-of-art

Recently, the accuracy of voice authentication system has increased significantly due to the successful application of the identity vector (i-vector) model. This paper proposes a new method for i-vector extraction. In the method, a perceptual wavelet packet transform (PWPT) is designed to convert speech utterances into wavelet entropy feature vectors, and a Convolutional Neural Network (CNN) is designed to estimate the frame posteriors of the wavelet entropy feature vectors. In the end, i-vector is extracted based on those frame posteriors. TIMIT and VoxCeleb speech corpus are used for experiments and the experimental results show that the proposed method can extract appropriate i-vector which reduces the equal error rate (EER) and improve the accuracy of voice authentication system in clean and noisy environment.

https://www.mdpi.com/1099-4300/20/8/600/pdf

Kun She

Papers

Identity Vector Extraction by Perceptual Wavelet Packet Entropy and Convolutional Neural Network for Voice Authentication.