Visual speech synthesis using statistical models of shape and appearance

B Theobald; JA Bangham; I Matthews; GC Cawley

Visual speech synthesis using statistical models of shape and appearance

B Theobald
, JA Bangham
, I Matthews
, GC Cawley

Research output: Contribution to conference › Paper

8 Citations (Scopus)

Abstract

In this paper we present preliminary results of work towards a video-realistic visual speech synthesizer based on statistical models of shape and appearance. A sequence of images corresponding to an utterance is formed by concatenation of synthesis units (in this case triphones) from a pre-recorded inventory. Initial work has concentrated on a compact representation of human faces, accommodating an extensive visual speech corpus without incurring excessive storage costs. The minimal set of control parameters of a combined appearance model is selected according to formal subjective testing. We also present two methods used to build statistical models that account for the perceptually important regions of the face.

Original language	English
Pages	78-83
Number of pages	6
Publication status	Published - Sept 2001
Event	International Conference on Auditory-Visual Speech Processing (AVSP-2001) - Aalborg, Denmark Duration: 7 Sept 2001 → 9 Sept 2001

Conference

Conference	International Conference on Auditory-Visual Speech Processing (AVSP-2001)
Country/Territory	Denmark
City	Aalborg
Period	7/09/01 → 9/09/01

Access to Document

http://www.isca-speech.org/archive_open/avsp01/av01_078.html

Cite this

@conference{c200e5b4235e40b486408662072267d6,

title = "Visual speech synthesis using statistical models of shape and appearance",

abstract = "In this paper we present preliminary results of work towards a video-realistic visual speech synthesizer based on statistical models of shape and appearance. A sequence of images corresponding to an utterance is formed by concatenation of synthesis units (in this case triphones) from a pre-recorded inventory. Initial work has concentrated on a compact representation of human faces, accommodating an extensive visual speech corpus without incurring excessive storage costs. The minimal set of control parameters of a combined appearance model is selected according to formal subjective testing. We also present two methods used to build statistical models that account for the perceptually important regions of the face.",

author = "B Theobald and JA Bangham and I Matthews and GC Cawley",

year = "2001",

month = sep,

language = "English",

pages = "78--83",

note = "International Conference on Auditory-Visual Speech Processing (AVSP-2001) ; Conference date: 07-09-2001 Through 09-09-2001",

}

TY - CONF

T1 - Visual speech synthesis using statistical models of shape and appearance

AU - Theobald, B

AU - Bangham, JA

AU - Matthews, I

AU - Cawley, GC

PY - 2001/9

Y1 - 2001/9

N2 - In this paper we present preliminary results of work towards a video-realistic visual speech synthesizer based on statistical models of shape and appearance. A sequence of images corresponding to an utterance is formed by concatenation of synthesis units (in this case triphones) from a pre-recorded inventory. Initial work has concentrated on a compact representation of human faces, accommodating an extensive visual speech corpus without incurring excessive storage costs. The minimal set of control parameters of a combined appearance model is selected according to formal subjective testing. We also present two methods used to build statistical models that account for the perceptually important regions of the face.

AB - In this paper we present preliminary results of work towards a video-realistic visual speech synthesizer based on statistical models of shape and appearance. A sequence of images corresponding to an utterance is formed by concatenation of synthesis units (in this case triphones) from a pre-recorded inventory. Initial work has concentrated on a compact representation of human faces, accommodating an extensive visual speech corpus without incurring excessive storage costs. The minimal set of control parameters of a combined appearance model is selected according to formal subjective testing. We also present two methods used to build statistical models that account for the perceptually important regions of the face.

M3 - Paper

SP - 78

EP - 83

T2 - International Conference on Auditory-Visual Speech Processing (AVSP-2001)

Y2 - 7 September 2001 through 9 September 2001

ER -