PCA method for automated detection of mispronounced words

Abstract

This paper presents a method for detecting mispronunciations with the aim of improving Computer Assisted Language Learning (CALL) tools used by foreign language learners. The algorithm is based on Principle Component Analysis (PCA). It is hierarchical with each successive step refining the estimate to classify the test word as being either mispronounced or correct. Preprocessing before detection, like normalization and time-scale modification, is implemented to guarantee uniformity of the feature vectors input to the detection system. The performance using various features including spectrograms and Mel-Frequency Cepstral Coefficients (MFCCs) are compared and evaluated. Best results were obtained using MFCCs, achieving up to 99% accuracy in word verification and 93% in native/non-native classification. Compared with Hidden Markov Models (HMMs) which are used pervasively in recognition application, this particular approach is computational efficient and effective when training data is limited.

Accent classification with phonetic vowel representation

“…Compared with Principle Component Analysis (PCA), which transforms data into eigenspace and preserves the data dimensions with larger variation [5], Linear Discriminant Analysis (LDA) reduces dimensions by mapping data into a subspace while maximizing the discriminative information. Assume there are K =…”

Improved accent classification combining phonetic vowels with acoustic features

“…PCA is commonly used for dimension reduction, which preserves the data dimensions with larger variations in the eigenspace. It has been applied to many applications, such as face recognition [13] and speech evaluation [14], etc. It also helps to regularize the data and avoid over-fitting in HLDA which is performed afterwards [15].…”