LIACS > Onderwijs > Master Computer Science > Masterprojecten
headerimg

Masterprojecten

Deze pagina is alleen beschikbaar in het Engels.

Here you can find an incomplete list of available Master Projects, sorted by thesis advisor.
Of course you can also pick your own subject and thesis advisor.
You can get to know more about projects and research groups during the 2nd year mandatory Master Class course.

LIACS Thesis Subjects

See this Google Group for a list of projects.

Farhad Arbab

Farhad Arbab

  1. Intuitionistic Temporal Linear Logic as a Transactional Coordination Language
    Intuitionistic Temporal Linear Logic (ITLL) is an extension of Linear Logic with temporal logic modalities. ITLL enables reasoning about epochs of time as well as resource usage. ITLL can also be used to encode the coordination language Reo and Zero-safe Petri nets, and can thus be seen as a transactional coordination language. In fact, it can express interaction patterns beyond those that Reo and Zero-safe Petri nets can express. The encoding may also result in more efficient implementations of Reo, based on the vast research on Linear Logic.

    The goal of this project then is to become familiar with ITLL and to implement a coordination language based on it, initially for the restricted fragment corresponding to Reo. The project will then consider extensions such as support for a more general notion of transaction.

  2. Evaluating Distributed Algorithms for Reo
    Recently, a number of approaches have been developed for implementing the coordination language Reo (centralized or decentralized, connector colouring or constraint automata based). Many more ideas and optimizations remain to be investigated (two phase commit, transaction-based). The goal of this project is twofold. Firstly, experiments will be designed to evaluate the costs and limitations of the various implementation techniques. Secondly, alternative implementation techniques and improvements will be explored.
  3. Translation of Creol into Scala Actors
    Creol is an active object based modelling language. Presently, it is implemented by translation into Maude rewriting framework, which limits the size of models that can be executed (though it does offer other benefits). In order to enable a more scalable implementation of Creol to enable experimentation with larger models, a more native implementation is required. As Scala's actors are very close in spirit to Creol active objects, this project tackles the problem of compiling Creol objects into Scala actors.
  4. Implement a testing framework for (Scala) actors.
    Complex networks of actors or active objects promise to offer a better approach to untilizing multicore processors than Java's threading model. As with any concurrent program, testing is quite challenging, often because buggy scenarios are difficult to reproduce. The nature of actors makes them much easier to test, because their state is totally encapsulated, so repeating the same series of message sends to an actor should produce equivalent results.

    The goal of this project is to design and implement a framework for testing actors (in the Scala language) in isolation, that is, extracted from the context in which they would run. Scenarios for testing can be generated based on logical expressions which describe the expected and undesirable behaviour of an actor, so part of this project will implement such techniques. As a case study, the testing framework developed in this project will be applied to the on-going actor-based implementation of Reo.

Nies Huijsmans

Nies Huijsmans

Computer Imagery project ideas 2007

  1. The translation of topological textual data into latitude-longitude numeric data and vice versa(open masters or project)
  2. Visualising data records by using associated location and time attributes (spatiotemporal view)(open masters or project)
  3. Interactive clustering in spatiotemporal dataviews (open masters or project)
  4. Creating 3D models of small archeological artifacts (with Paul Boon)
  5. Clustering similar result pages in search engine output: for instance those in a meta book search engine like www.addall.com by creating a representative entry for each group.
  6. Analysis and visualisation of web server log data: how to effectively use log data of server supported dynamic sites in optimising the way the site functions.
  7. Piece-wise continuous image theory: sound mathematical foundation of image processing functions.
  8. 3D reconstruction from (almost) crossing directions using conformal geometric algebra and a volumetric model.
  9. life size stereo projection and spatial interaction with the past: depth map from stereo pair is used to create 3D model that is combined with momentary images of a person wandering around in this (past) scene.
  10. Automatic positioning and annotation of family pictures: to bridge the semantic gap; portrait/landscape; up/down; how many people; man/woman; old/young etc.
  11. VIS "Image Submarine": Visual Image Searching with under-water annotation: how to effectively use textual annotation with only visual interaction.
  12. Imaging Events: extending geotag -> events: geographic location,time,person,object, detecting object/person events in time-space data (like biographies) and how to visually support these events.
  13. From videosequence to 3D model: given a videosequence taken around a 3D object by first finding the trajectory followed by the camera.

Prof.dr. T. Baeck

Natural Computing group,

see Natural Computing website.

 

 

Last edited on 27/07/2010 2:53pm