CMVS kandidaatintyön aiheet

 

Tällä sivulla esitellään Konenäön ja signaalianalyysin tutkimuskeskuksen (Center for Machine Vision and Signal Analysis) tarjoamia kandidaatintyön aiheita. Kandidaatintyö voidaan tehdä 1-4 opiskelijan ryhmissä ja kaikkien aiheiden laajuus sovitetaan opiskelijoiden taustan ja lukumäärän mukaan. Kandidaatintyön dokumentointikieli on pääsääntöisesti suomi. Tarkemmat ohjeet kandidaatintyön tekemiseksi löytyvät tietotekniikan sivuilta.

 

Tarjotut aiheet ja yhteyshenkilöt:

Datankeruu-plugin Crome-selaimelle

Video recorder plug-in for WebRTC on Crome browser. The major objective of this topic is to provide for a questionnaire tool that enables the researchers of facial expressions analysis to obtain test material for research purposes. Both sound and video should be recorded. The project should likely be developed based on Kurento. The student makes the background study and informed decisions about the platform. Contact person: Olli Silvén

Kuvapohjainen pullonpalautusjärjestelmä

A Molok-type reverse vending imaging system for cans and plastic bottles. The main idea is to capture images of cans and bottles dropped to a container through an opening. The opening is equipped with 3-4 cameras and perhaps stroboscopic illumination. The minimum target level in this topic is the capability to read bar codes.​ In the other end of the scale advanced image analysis and 3D reconstruction could be made based on SURF-features. Contact person: Olli Silvén

Kasvojenilmeiden analysointialgoritmien verkkoselainintegraatio

This topic focuses on reimplementation of facial expressions analysis algorithms for web-browser integration. The project should likely be developed based on Kurento. The student makes the background study and informed decisions about the implementation technique. Contact person: Olli Silvén

Langaton hintalappu (kirjallisuuskatsaus)

Wireless price tag technology background and foresight. Key questions to consider include: How have the current wireless price tags with displays been implemented? How should they be implemented in the future? Which existing technology components are relevant (e.g., 4G LTE PSM?). What is the role of positioning in price tag logistics? Imaging enhanced price tags?​ The student will perform a literature review that in this topic borders both technology and market surveys. Contact person: Olli Silvén

Koneoppimisen soveltaminen käytännön ongelmiin

Various topics where machine learning is applied to simple problems using for example Matlab CNN Toolbox or Caffe Toolkit are offered. Potential problems include (but are not limited to) for example labeling of human body parts and recognizing hand gestures from images or videos. The data sets will be provided. Contact persons: Vili Kellokumpu, Esa Rahtu

​Kuvankäsittelyn projekti

Project on image processing. Various image processing related topics are offered and tailored based on the interests and experience of the students. Contact person: Janne Heikkilä

Aivosähkökäyrien (EEG) analyysi LORETA-työkalulla

Analyzing electroencephalograms (EEGs) with LORETA (Low Resolution Electromagnetic Tomography) software. The student will utilize the online available software package LORETA for the functional localization and/or connectivity analysis of EEGs recorded e.g. during anesthesia or from patient with brain injury. Contact person: Jukka Kortelainen

Toiminnallisen kytkennällisyyden mittarit aivosähkökäyristä (kirjallisuuskatsaus)

Functional connectivity measures used in the EEG analysis. The student will perform an analytical literature review on the state-of-the-art methods used when studying the functional connectivity of the brain from EEG data. Contact person: Jukka Kortelainen

Aivojen stimuloinnin aikaansaamat aivosähkökäyrien ilmiöt (kirjallisuuskatsaus)

EEG phenomena induced by brain stimulation. The student will perform an analytical literature review on the EEG phenomena which can be induced with electrical and/or magnetic brain stimulation. Contact person: Jukka Kortelainen

Fysiologisten signaalien analysointi lääketieteellisissä sovelluksissa

Various topics on physiological signal processing for medical applications using methods ranging from image and signal processing to machine learning are offered. Contact person: Tapio Seppänen

Oma aihe

Mikäli sinulla on mielessä oma aihe kandidaatintyölle, voit ottaa suoraan yhteyttä henkilöön, jonka asiantuntemus sopii parhaiten aiheeseesi

Viimeksi päivitetty: 14.4.2016