Back to results
US 6,208,356 B1Granted
Image synthesis
AssigneeBritish Telecommunications public limited company
InventorsAndrew P Breen, Emma J Bowers
Life of this patent
Estimating the term…
Abstract
Synthesis of a moving picture of a face (e.g. to accompany synthetic speech) is performed by converting an input phoneme string into a sequence of mouth shapes or visemes. Specifically a shape is generated for each vowel and for each transition involving a consonant.
Claims14
A method of generating signals representing a moving picture of a face having visible articulation matching a spoken utterance, said method comprising:
- receiving a sequence of phonetic representations corresponding to successive portions of the utterance;
- identifying a sequence of mouth shapes in accordance with said sequence of representations;
- generating for each identified mouth shape a command specifying that shape;
- generating intermediate commands each of which specifies a shape intermediate the shapes specified by the preceding and following commands;
- said representations comprising a first set of representations that represent portions of a first phonetic kind, and a second set of representations that represent portions of a second phonetic kind;
- said identification of a sequence of mouth shapes including:
- identifying a mouth shape for each representation from said first set;
- and
- identifying a mouth shape for each transition from a phonetic representation in said first set to a phonetic representation in said second set, for each transition from a phonetic representation in said second set to a phonetic representation in said first set and for each transition from a phonetic representation in said second set to a phonetic representation in said second set.
6 more elements
An apparatus for generating signals representing a moving picture of a face having visible articulation matching a spoken utterance, said apparatus comprising:
- means arranged in operation to receive a sequence of phonetic representations corresponding to successive portions of the utterance and in response thereto to
- identify a sequence of mouth shapes in accordance with said sequence of phonetic representations;
- generate for each identified mouth shape a command specifying that shape;
- generate intermediate commands each of which specifies a shape intermediate the shapes specified by the preceding and following commands;
- said representations comprising a first set of representations that represent portions of a first phonetic kind, and a second set of representations that represent portions of a second phonetic kind;
- means to identify a mouth shape for each phonetic representation in said first set;
- and
- means to identify a mouth shape for each transition from a phonetic representation in said first set to a phonetic representation in said second set, for each transition from a phonetic representation in said second set to a phonetic representation in said first set and for each transition from a phonetic representation in said second set to a phonetic representation in said second set.
5 more elements
Dependencies are inferred from the claim text.