Image Processing Reference
In-Depth Information
0.94 A quarter-ounce of chocolate = four pounds of fat.
(Slick's Second Universe Law)
0.82 Equal amounts of dark chocolate and white chocolate,
is a balanced diet.
User: What can you say about children?
Alice: I will think about it and get back to you.
EHeBby:
0.68 Praise your children openly, reprove them secretly.
(W. Cecil)
0.68 When speaking to your children: say what you mean,
mean what you say,mbut don't be mean when you say it.
User: What do you think of life?
Alice: Life is the opposite of death.
EHeBby:
0.987 Life is an echo.
0.986 Life is a roller coaster.
User: Have you ever seen Santa Claus?
Alice: No I haven't.
EHeBby:
0.983 The Bermuda Triangle got tired of warm weather.
It moved to Finland. Now Santa Claus is missing.
0.746 There are 4 Santa Claus stages:1) When you believe in Santa.
2) When you donŠt believe in Santa.3) When you are Santa.
4) When you look like Santa.ve not seen it. What's it like?
7. CONCLUSION
A complete framework for an emotional talking head able to manage humor while conversing
with the user has been presented along with its implementation. The whole architecture relies
on a suitable AIML-based chatbot, and an animation engine for the talking head.
The chatbot reasoner module is based on an extended AIML architecture where both humor,
and emotions can be dealt with using suitable tags. A computational humor engine is able
both to detect and to generate humorous sentences. Humor detection relies on the presence
of alliteration, antinomy, or adult slang in the user's utterances, which are searched for using
suitable thesauri like CMU and WordNet. Generation of humor makes use of a LSA based
semantic space where humorous sentences have been placed along with the conversation
topics.
The system can also select the control parameters for the animation engine, regarding the
mesh deformation due to the emotive state to be expressed, the prosody for controlling speech
generation, and the coarticulation model that is used to morph a set of key visemes related to
phonemes.
The whole system has been tested on the humor recognition task with satisfactory results.
However, our system is currently under development and much work has to be done in order
to improve the whole architecture. Humor recognition algorithms can be enhanced, in order
to capture different grades of humor, and to fully exploit the different levels of intensity in
Talking Head emotional expressions.
The emotion database has to be completed al least with all the six Ekman basic emotions.
Moreover, the most recent emotion models Ekman (1999) use more than six basis emotional
Search WWH ::




Custom Search