Acoustic diagnostics – Defect detection | Signal evaluation | Quality assurance


Classification method See ▶ pattern recognition
Deep learning ▶ Machine learning method for deep neural networks
Deep neural network (DNN) Artificial neural network with numerous hidden layers
▶ pattern recognizer for ▶ (sequences of) feature vectors
EM algorithm ▶ Machine learning method for ▶ Gaussian mixture models and
▶ hidden Markov models
Feature analysis Method for calculation of ▶ (sequences of) feature vectors from measured signals
Feature vector (sequence) Set of ▶ classification-relevant numerical parameters, possibly as a temporal sequence
Gaussian mixture model (GMM) Statistical ▶ model for ▶ pattern recognition in ▶ feature vectors based on Gaussian mixture distribution densities
Hidden Markov model (HMM) Statistical ▶ model for ▶ pattern recognition in ▶ feature vector sequences based on a Markov process, e.g., GMM-HMM, DNN-HMM
Machine learning Automated method for building of ▶ models for
▶ pattern recognition and decision-making processes
Model Here: computational representation of knowledge
Pattern recognition Method for differentiation into predefined classes, e.g., based on ▶ DNN, ▶ GMM, ▶ HMM, or ▶ SVM
Primary analysis First step in ▶ feature analysis (signal processing, e.g., filter banks, FFT, STFT, DWT, Cepstrum, LPC, Wigner-Ville distribution, etc.)
Secondary analysis Second step in ▶ feature analysis (statistics, data compression, e.g., quantiles, moments, differences, filtering, PCA, LDA, ICA, JFA, etc.)
Semantic processing Computational processing of meanings (e.g., of measured signals)
Sequence classifier ▶ Pattern recognizer for a sequence of ▶ feature vectors
Signal analysis See ▶ primary analysis
Support vector machine (SVM) ▶ Pattern recognizer for ▶ feature vectors
Training method See ▶ machine learning
Vector classifier ▶ Pattern recognizer for a ▶ feature vector