Master’s Thesis: 2D-3D Data Fusion in Interior (Re-) Design
Institute DIGITAL
Place of employment: Graz
Start: to be agreed on
Time limit: 6 months
Extent of working hours: Part time, up to 50% (19,25 hours per week)
Payment: 796,50 gross for 50% employment, payable 14 times per year (aliquot)
In our research group Machine Vision Applications we are awarding, in cooperation with an Austrian university, the following Master Project / Master’s Thesis:
2D-3D Data Fusion in Interior (Re-) Design
Description:
Automatic room modeling is a growing field in interior architecture, planning and real estate marketing. The strategy of 3D modeling combined with AI-based 2D and 3D segmentation depends on the employed sensor model, with different levels of density, accuracy, coverage and data entities. For a collaborative project between industry, research and university, we are looking for a motivated student for developing methods and data structures to foster the fusion of 2D (image) and 3D (point cloud) data, aiming at interior models that combine the benefits of the used data sources.
Objectives:
- Incorporating single images / panoramas, image poses and point clouds from state-of-art (commercial) interior capturing sensors
- Incorporate 2D interior segmentation tools existing at JR
- Incorporate 3D room segmentation tools existing at JR
- Investigate and implement the interconnections between the 2D segmentation and 3D modeling data structures and algorithms for an optimized representation
- Work towards a 2½ or 3D room representation, complement to or based on current industry standards, to enable the CAD-based planning & construction of built-in furniture
- Enable the ergonomic visualization of intermediate processing products as well as the interconnections between 2D and 3D data structures to allow industrial users to assess the validity of the process
Qualifications:
- High commitment to the project work
- Computer vision and/or 3D vision & visualization skills
- Interest in 3D room modeling and AI-based room segmentation
- Interest in data structures
- Programming skills, preferably C++ & Python
What we offer:
- minimum monthly salary according to the applicable collective agreement € 1,593 gross, based on a full-time position
- a team-oriented working atmosphere and the most modern infrastructure
- flexible working hours with the possibility of teleworking
- targeted support for your personal career through the internal training program and in the further development of your professional expertise
- additional full days off on Good Friday, 24.12. and 31.12.
- attractive social benefits such as sports and culture promotion, mobility allowance, company events, free coffee, etc.
We look forward to receiving your meaningful application and ask you to send it electronically to the following address:
JOANNEUM RESEARCH Forschungsgesellschaft mbH
email:PEMBewerbungen@joanneum.at
Subject: DIGITAL Master Thesis 2D-3D Data Fusion
Contact person: Gerhard Paar