Last edition winners
-
1
SAFE — Speech Alignment and Feature Extractor
Leônidas Silva Jr.
SAFE (Speech Alignment and Feature Extractor) is a fully automatic protocol (i.e., a system) that performs file renaming, a two-step phonetic-acoustic alignment of speech/text file pairs, and the extraction of prosodic-acoustic and phonetic-segmental measures. The entire protocol runs automatically from start to finish in an end-to-end workflow. This study represents the current state of the art and seeks to provide the Brazilian and international scientific community with a robust and efficient tool that can be applied to research across several areas of the phonetic sciences, including Linguistics, Clinical research, Forensic science, Computer Science, Artificial Intelligence (AI), and Entertainment, among others.