Terminology...The Learning Curve
- Yavneeka Patel
- Mar 11, 2015
- 1 min read
In attempting the speech recognition portion, I'm rapidly discovering that current documentation for these programs assume you have quite a bit of background information. Of which, I have about 1% knowledge. So I've started this list to help me keep track of concepts and terminology...
Speech:
PHONES: Similar classes of sound.
DIPHONES: Parts of phones between two consecutive phones. These can be vowels or consonants placed next to each other. The Spanish language has about 800 diphones.
TRIPHONES: Phones that describe specific sounds. Unlike diphones, they are matched with the same range in waveform as just phones.
SENOMES: Small amount of distinct short sound detectors that detect variety in sound. Sphynx uses 4000 distinct short sound detectors to compose detectors for triphones.
FILLERS: Non-linguisting sounds and words (breathing, um, uh)
UTTERANCES: Breaks in speech made by Fillers.
HIDDEN MARKOV MODEL:
Maven:
MIRROR: Where files are downloaded from on the internet. It is a http link. [Seen on Maven download site]
Recent Posts
See AllSo we've made it to the very end. Through all the ups and downs and we're all still standing. It was an amazing semester and I made some...
/** *Java Speech Grammar Format (JSGF) file *This file holds all the words that the program will recognize **/...
/* * Copyright 2013 Carnegie Mellon University. * Portions Copyright 2004 Sun Microsystems, Inc. * Portions Copyright 2004 Mitsubishi...
Comments