A tentative effort at grouping the interest areas of the COST 249 partners is given below.
a) Dialogue
i) Confidence measures -dialogue
ii) Dialogue strategies - inquiry systems
b) Assessment, robustness
i) SpeechDat Mobile
ii) SpeechDat Fixed Net
c) Applications, design
i) Voice dialling
ii) Dictation systems
iii) Application integration
iv) Assessment for user-oriented design
v) WWW-based applications
vi) Human factors
vii) Application of confidence measures
a) Hybrid statistical and knowledge-based approaches
b) Language models
c) Pronunciation modelling
d) Acoustical, semental units
e) Large vocabulary recognition
f) Morpheme based ASR
g) Word spotting
a) Robustness
i) Robust acoustic processing
ii) Local distances
iii) Multi-band approaches
iv) Multiple feature sets
b) Features
i) Articulatory based ASR
ii) Prosodic features
iii) Non-linear acoustic modelling
c) Databases - design, collection and annotation
d) Adaption