Feature extraction for speech recognition based on orthogonal acoustic-feature planes and LDA

This paper describes an attempt to extract multiple topological structures, hidden in time-spectrum (TS) patterns, by using multiple mapping operators, and to incorporate the operators into the feature extractor of a speech recognition system. In the previous work, the author proposed a novel feature extraction method based on MAFP/KLT (MAFP: multiple acoustic-feature planes), in which 3×3 derivative filters were used for mapping operators, and showed that the method achieved significant improvement in preliminary experiments. In this paper, firstly, the mapping operators are directly extracted in the form of a 3×3 orthogonal basis from a speech database. Next, the operators are evaluated, together with 3×3 simplified operators modeled on the orthogonal basis. Finally, after comparing the experimental results, the author proposes an effective feature extraction method based on MAFP/LDA, in which a Sobel filter is used for mapping operators