Francois Brémond

Motivated by the growing of ageing population worldwide and the need to concentrate research efforts on a specific target group, our research focuses on ageing persons with physical and cognitive deficiencies. The primary goal is to enable the person with mild dementia, through assistive technologies, to maximize his physical and mental function, and to continue to engage in social networks, so that he can continue to lead an independent and purposeful life.

The person with mild dementia usually has little problems in the actual performance of tasks in most basic activities of daily living but is often handicapped by his poor memory and thus forgets to carry out these tasks. These can include bathing, changing clothes and taking medication on time. Assistive technology that provides timely prompts and reminders will enable him to preserve his abilities and independence.

Finding an appropriate interface to interact with persons with dementia can be a challenge. Indeed, the interaction should be done through a user-friendly interface by making it easier to access the environment and for the person to benefit from assistance. Introducing a desktop computer with a keyboard is not always well accepted. Thus, we have adopted the approach of providing assistive service through a multimodal interactive system including TV, iPad-like tablets and wireless speakers. Consequently, the reasoning level of the system may be calibrated according to the contextual situation of the user (context awareness) to bring about the service required. The User Interface should support not only the user’s preferences but also his profile as dictated by his cognitive abilities. At the same time, different surrounding computing platforms and devices/sensors are considered additional sources of information. In this paper we will focus mainly on the interaction level with the system as well as on the validation stages performed to meet users’ requirements in terms of user interface design and content of service. This is the result of more than 3 years work, from 2006 to 2010, within 2 Europeans projects (IST-FP6 Cogknow project and ITEA NUADU project) and an ongoing project in Singapore (AMUPADH).

Tracking multiple objects under merging, splitting, and occlusion situations is a challenging task for a surveillance system, especially when objects are close together. This is due to ambiguity and uncertainty of visual features from these objects as compared to a situation where tracked objects are isolated. To handle these challenges in multiple object tracking, techniques that apply observations from detected objects have been introduced. Joint probabilistic data association (JPDA) are one of them and calculate the probability of all possible data association between objects and observations. The advantages of JPDA and particle filter can then be combined to obtain better tracking results from human detection algorithm. The detection stage can be performed using Local Binary Pattern (LBP). LBP is a texture descriptor that combines local primitives into a feature histogram. LBP and its extensions outperform existing texture descriptors both with respect to performance and to computational efficiency. It is then suitable for foreground detection or background modelling.

Most methods for multiple camera tracking also rely on accurate calibration to associate data from multiple cameras. However, it is often not easy to have an accurate calibration in some real applications due to practical reasons. The inaccurate calibration can then lead to wrong data association of objects between cameras. To handle the data association of objects in multiple cameras under inaccurate ground plane homography, the RFS Bayes filter can be applied. Observations measurements can be modelled from cameras to a random finite set. This random finite set includes the primary measurement from the object, extraneous measurements of the object, and clutter. Experimental results including challenging cases such as occlusions and merging persons will be presented.