Articulated motion discovery using pairs of trajectories

We propose an unsupervised approach for discovering characteristic motion patterns in videos of highly artic-ulated objects performing natural, unscripted behaviors, such as tigers in the wild. We discover consistent patterns in a bottom-up manner by analyzing the relative displace-ments of large numbers of ordered trajectory pairs through time, such that each trajectory is attached to a different moving part on the object. The pairs of trajectories de-scriptor relies entirely on motion and is more discriminative than state-of-the-art features that employ single trajecto-ries. Our method generates temporal video intervals, each automatically trimmed to one instance of the discovered behavior, and clusters them by type (e.g., running, turn-ing head, drinking water). We present experiments on two datasets: dogs from the YouTube objects and a new dataset of National Geographic tiger videos. Results confirm that our proposed descriptor outperforms existing appearance-and trajectory-based descriptors (e.g., HOG and IDTF) on both datasets and enables us to segment unconstrained an-imal video into intervals containing single behaviors. 1