Finally, we want to evaluate the empirical performance and scalability of our approach under a set of controlled experiments. The resulting beliefs can also be easily accessed and retrieved by the other subarchitectures. Finally, Section 8 concludes this report, and provides directions for future work. The outcome is a set of possible unions, each of which has an existence probability describing the likelihood of the grouping. Multivariate probability distributions over possible values are used to account for the partial observability of the data, while the first-order expressivity of Markov Logic allows us to consisely describe and reason over complex relational structures. The feature values can be either discrete as for categorical knowl- edge or continuous as for real-valued measures.

The code has recently been fully refactored and includes an easy-to-use graphical interface. This is realised by checking whether their respective features correlate with each other. Sen- sory data is extracted and processed by the sensory subarchitecture s. As such, it provides an elegant account of both the uncertainty and complex- ity of situated human-robot interactions. Such perceptual belief i would be formally defined as: Section 5 then explains step-by-step how such beliefs can be constructed bottom-up, iteratively, from perceptual inputs. Spoken language interaction with model uncertainty:

pierre lison thesis

Continual processing of thesos dialogue in human-robot collaborative activities. A system for continuous learning of visual concepts. Shared beliefs contain information which is part of the common ground for the group [3]. We leave this question as an interesting topic for future research.

Pierre Lison Completes Doctoral Degree – Department of Informatics

Its main purpose is to serve as the voice of young researchers in the public arena, through a range of initiatives related to research policy and science dissemination. Computer Speech and Language, 21 2: Beliefs evolve dynamically over time, thsis are constructed by a three-fold iterative process of information fusion, refinement and abstraction.


Anal- ogous to perceptual grouping which seeks to bind observations over modalities, tracking seeks to bind beliefs over time. I also participated in the centre for research-based innovation Big Insight: In Perception and Interactive Technologies: The advantage of using such representation is the possibility of using logical inference mechanisms to reason over such structure and automatically derive new beliefs.

Proceedings of the 21st national conference on Artificial intelligence, pages — I am delighted to visit the Idiap research institute in Switzerland for a period of 3 months from August to October in order to work together with Andrei Popescu-Belis on dialogue modelling for statistical machine translation.

Remember me on this computer. Section 2 provides a brief intro- duction to Markov Logic, the framework used for belief modelling. Proceedings of the 23rd national conference on Artificial intelligence, pages — The first public release of the OpenDial toolkit is finally available!

Belief models are usually expressed as high level symbolic representations merging and abstracting information over multiple modalities. Pierre Lison holds a M. This in- cludes both the evolution of the physical environment, and the evolution of the interaction itself.

PhD thesis, MIT, Graphical models in a nutshell. Skip to main content.


pierre lison thesis

Information fusion refers to the operation of merging data from distinct knowledge sources into one single representation. Context in abductive interpretation.

Pierre Lison Completes Doctoral Degree

Now, off to a week of well-deserved vacation ;- [] A-magasinet featured this week a three-pages article on Lenny and pierr research on human-robot interaction through spoken dialogue. Statistics for the knowledge economy.

The architectural schema is based on a distributed set of subarchitectures. Perceptual beliefs are inserted onto the binder by the various subsystems included in the architecture. Let us pieree for our example two uncertainties: Finally, Section 8 concludes this report, and provides directions for future work.

New Research Scientists have joined SAMBA | Norsk Regnesentral

Moreover, the spatio-temporal frame can be extended with the notion of per- spective, where spatial and temporal constraints are defined as being relative to a particular agent a. Besides the implementation, future work will focus on three aspects.

Multi-modal fusion is also specified in a Markov Logic Tjesis. The epistemic status of this information is attributed.

OpenDial is a Java-based toolkit to develop robust and adaptive dialogue systems for various domains. They can also be used by perceptual pierre to adapt their internal processing operations to the current situated context contextual priming, anticipation, etc.

pierre lison thesis