marianne the representing the school of and electronic electrical and computer engineering this afternoon and
the title of head tool it's
towards better automated for systems
modeling accents for automatic
two recognition
i want you to remember all those times you had to use an automated phone
system
how was your experience how did you feel
a bit you feel frustrated by don't number of times you have to repeat which
is that older an over again
even then you might have been transferred to the wrong department
but no that's is over you have to wait and then long q listening to
a web had to solve and a voice that
you're called is extremely important to us ulysses us assume that the next agencies available
despite all this stuff estimation that we go through using these speech enabled automated phone
systems rather than the traditional types to one
the save an average cost and set up to three million pounds manually
by two thousand and seventeen this industry could be more up to two billion pounds
after all there seems to be no way to avoid you it's speech enabled voice
jails
but why and these systems understandable be stated
and they are not that clever human
they cannot adapt then you speakers back then
in fact
a few months ago you mean intensity con still had to stand eleven million pounds
an automated phone system
that we couldn't work to be the local accent
the goal of my phd just all these you mean and i can't robust automatic
speech recognition system
so
i have this massive database of recorded speech from people be different regional accent
then an unknown speaker talks to my system it recognises their accent it then selects
the data from a there is speakers with similar accent in order to create a
more personalised model out there is speech
using the is really improve the performance of my system incredibly
in fact
the error rate will be reduced by up to fifty path and for scientific contact
them
such as
lexington
you can't result of this is that customers
we have more productive time to use during their day instead of a tuning call
center cues and how companies
thus have cu were number of angry customers
in the next page of my clear i would like to address the issue space
by my now ready to fight as children elderly people with the speech this so
the thank you