Unsupervised Learning of Shape-Motion Patterns for Objects in Urban Street Scenes

Dirk Klostermann, Aljosa Osep, Jörg Stückler and Bastian Leibe

Abstract

Tracking in urban street scenes is predominantly based on pretrained object-specific detectors and Kalman-filter based tracking. More recently, methods have been proposed that track objects by modelling their shape as well as ones that predict the motion of objects using learned trajectory models. In this paper, we combine these ideas and propose shape-motion patterns (SMPs) that incorporate shape as well as motion to model a variety of objects in an unsupervised way. By using shape, our method can acquire trajectory models that distinguish object categories with distinct behaviour. We develop methods to classify objects into SMPs and to predict future motion. In experiments, we analyse our learned categorization and demonstrate superior performance of our motion predictions compared to a Kalman-filter and a learned pure trajectory model. We also demonstrate how SMPs can indicate potentially harmful situations in traffic scenarios.

Session

Motion and Tracking

Files

Extended Abstract (PDF, 290K)
Paper (PDF, 2M)
Supplemental Materials (ZIP, 19M)

DOI

10.5244/C.30.68
https://dx.doi.org/10.5244/C.30.68

Citation

Bibtex

        @inproceedings{BMVC2016_68,
        	title={Unsupervised Learning of Shape-Motion Patterns for Objects in Urban Street Scenes},
            author={Dirk Klostermann, Aljosa Osep, J{\"o}rg St{\"u}ckler and Bastian Leibe},
        	year={2016},
        	month={September},
        	pages={68.1-68.12},
        	articleno={68},
        	numpages={12},
        	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
        	publisher={BMVA Press},
        	editor={Richard C. Wilson, Edwin R. Hancock and William A. P. Smith},
        	doi={10.5244/C.30.68},
        	isbn={1-901725-59-6},
        	url={https://dx.doi.org/10.5244/C.30.68}
        }