The LaMIT database: a read speech corpus for acoustic studies of the Italian language toward lexical access based on the detection of landmarks and other acoustic cues to features
The LaMIT database consists in recordings of 100 Italian sentences. The sentences in the database were designed so to include all phonemes of the Italian language, and also take into account the typical frequency of each phoneme in written Italian. Four native adult speakers of Standard Italian, raised and living in Rome, Italy, two female and two male, pronounced the sentences in two different recording sessions; two repetitions for each sentence per speaker were therefore collected, for a total of 800 recordings. The database was specifically created for application in the LaMIT project, that focuses on the application to the Italian language of the Lexical Access model proposed by Ken Stevens for American English. The model relies on the detection of specific acoustic discontinuities – called landmarks – and cues correlated to acoustic features that characterize each phoneme. Each recording was thus processed to generate a set of labeling files that identify both predicted landmarks and cues, based on the phonemes in the written sentence and actual landmarks/cues detected in the recording. The labeling files, redacted according to the labeling syntax used in the Praat speech processing software, are also made available as part of the LAMIT database.