PGR meeting and Research Presentations – July 2016

The monthly PGR meeting was held on Wednesday 13th July, 14:00-16:00, Room MC3108.

We had 2 speakers for this month seminar:

* Claudio Coppola

We investigate how incremental learning of long-term human activity patterns improves the accuracy of activity classification over time.Rather than trying to improve the classification methods themselves, we assume that they can take into account prior probabilities of activities occurring at a particular time and location.We use the classification results to build spatial and temporal models that can provide these priors to the classifiers.As our system gradually learns about typical patterns of human activities, the accuracy of activity classification improves, which results in even more accurate priors. Two datasets collected over several months containing hand-annotated activity in residential and office environments were chosen to evaluate the approach.Several types of spatial and temporal models were evaluated for each of these datasets.The results indicate that incremental learning of daily routines leads to a significant improvement in activity classification.


* Evangelia Kotsiliti

Diabetic retinopathy is a complication of diabetes affecting the eye and a leading cause of blindness worldwide. In many countries around the world, systematic screening for diabetic retinopathy is provided to the patients diagnosed with diabetes in order to reduce the burden of blindness. However, considering the rising numbers of people who are diagnosed with diabetes every year, it is plausible to think that the provision of screening services to an increased demand may no longer be affordable. This projects aims at the utilisation of simple, commonly available patient characteristics and biochemical measures to identify patients at high risk of having retinopathy at the time of screening. A subsequent step aims at the development of a cost-effectiveness model to compare the cost and consequences of the risk model against the current manual grading. The ultimate outcome would be the development of reliable clinical model that can reduce the overall screening cost and retain effectiveness.


There will be no seminar in August. The date and venue for the next meeting in September 2016 will be announced.