Applied Informatics Group

Using spinImages for 3D object classification

The subject of this thesis is to evaluate if a combination of spin images and Bag of Features classifi cation can be used to categorize point clouds representing real world objects. At first, for each point cloud spin images are created as descriptors for the cloud's points. Then vector quantization is used to learn prototypes in the spin image space. These prototypes are then used as a codebook for extracting features for each point cloud in the training set, which means, that standard bag of features approach is used. Afterwards these features are used to train a support vector machine. Using test sets disjunct from the training set, best parameters for the SVM as well as for spin image generation will be computed. Two different datasets will be used, to evaluate the approach. One of these consists of handmade 3d models as found in a subset of the Princeton Shape Benchmark and one consists of 3D scans obtained by using a range scanner on real objects. Both used datasets consist of models which represent furniture of different kinds.
AttachmentSize
ThesisUnofficialCorrection.pdf2.01 MB

Recent Best Paper/Poster Awards

Goal Babbling of Acoustic-Articulatory Models with Adaptive Exploration Noise
Philippsen A, Reinhart F, Wrede B (2016)
International Conference on Development and Learning and on Epigenetic Robotics (ICDL-EpiRob) 
PUB | PDF

 

Are you talking to me? Improving the robustness of dialogue systems in a multi party HRI scenario by incorporating gaze direction and lip movement of attendees
Richter V, Carlmeyer B, Lier F, Meyer zu Borgsen S, Kummert F, Wachsmuth S, Wrede B (2016)
International Conference on Human-agent Interaction (HAI) 

 

"Look at Me!": Self-Interruptions as Attention Booster?
Carlmeyer B, Schlangen D, Wrede B (2016)
International Conference on Human Agent Interaction (HAI)
PUB | DOI

 

For members