Introduction to Speech Analysis with FAVE (Forced Alignment & Vowel Extraction)

Date:

17/02/2015

Organised by:

Scottish Graduate School of Social Science, SAIL Pathway

Presenter:

Dr Lauren Hall-Lew

Level:

Advanced (specialised prior knowledge)

Contact:

Dr Lauren Hall-Lew, Lauren.Hall-Lew@ed.ac.uk

Map:

View in Google Maps  (EH8 9AD)

Venue:

Charles Street
Dugald Stewart Building, University of Edinburgh

Description:

This one-day workshop will teach you how to use the computational tool FAVE: Forced Alignment and Vowel Extraction. The tool takes a sound file of speech and a word-level transcript of that speech and uses a speech recognizer to match the words in the transcript to the acoustic signal of the speech file, resulting in the automatic generation of a new text file that indicates all the time points for the start and end of every vowel, consonant, noise, and pause, all of which are automatically labelled. A researcher can then use that to automatically extract information about any set of sounds that might be of interest. For example, FAVE also includes a tool to automatically extract the key acoustic measures for vowels, allowing a dialectologist, sociolinguist, forensic linguist, or phonetician to easily quantify accents, dialects, and speaking styles.

This is an essential skill for any current postgraduate student in the aforementioned areas of speech research. However, although there is a web-based interface for the tool, the benefits of using it are very limited compared to the freedom and flexibility a researcher can have by installing the program on their own computer and understanding more about how it functions. This particular Advanced Training event is ideal for all students because the researcher leading the training Dr Fruehwald, is one of the original designers and continuing developers of FAVE.

The day will be comprised of a morning lecture introducing the benefits of learning how to use FAVE, followed by an afternoon clinic helping students learn how to use the program on their own laptops (students will need to bring individual laptops to participate in the event). Students who attend should know the basics about speech science, phonetics, and sociolinguistics.

Cost:

Free to attend but students from Universities outwith Scotland will need to pay for travel and accommodation themselves.

Website and registration:

Region:

Scotland

Keywords:

Behavioural Research, Hypothesis testing research, Data Collection, Quality in Quantitative Research, Measurement Error, Quantitative Software, Advanced Technologies , Measurement

Related publications and presentations:

Behavioural Research
Hypothesis testing research
Data Collection
Quality in Quantitative Research
Measurement Error
Quantitative Software

Back to archive...