Unsupervised Tracking From Clustered Graph Patterns

Fabien Diot (1, 2) , Elisa Fromont (1) , Baptiste Jeudy (1) , Emmanuel Marilly (2) , Olivier Martinot (2)

1 LHC - Laboratoire Hubert Curien
2 Alcatel-Lucent Bell Labs France [Nozay]

This paper shows how data mining and in particular graph mining and clustering can help to tackle difficult tracking problems such as tracking possibly multiple objects in a video with a moving camera and without any contextual information on the objects to track. Starting from different segmentations of the video frames (dynamic and non dynamic ones), we extract frequent subgraph patterns to create spatiotemporal patterns that may correspond to interesting objects to track. We then cluster the obtained spatio-temporal patterns to get longer and more robust tracks along the video. We compare our tracking method called TRAP to two state-of-the-art tracking ones and show on three synthetic and real videos that our method is effective in this difficult context.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Fabien Diot, Elisa Fromont, Baptiste Jeudy, Emmanuel Marilly, Olivier Martinot. Unsupervised Tracking From Clustered Graph Patterns. International Conference on Pattern Recognition, Aug 2014, Stockholm, France. 6 p. ⟨hal-00968304⟩