Timezone: »

Learning Probabilistic Non-Linear Latent Variable Models for Tracking Complex Activities
Angela Yao · Juergen Gall · Luc V Gool · Raquel Urtasun

Tue Dec 13 08:45 AM -- 02:59 PM (PST) @

A common approach for handling the complexity and inherent ambiguities of 3D human pose estimation is to use pose priors learned from training data. Existing approaches however, are either too simplistic (linear), too complex to learn, or can only learn latent spaces from ""simple data"", i.e., single activities such as walking or running. In this paper, we present an efficient stochastic gradient descent algorithm that is able to learn probabilistic non-linear latent spaces composed of multiple activities. Furthermore, we derive an incremental algorithm for the online setting which can update the latent space without extensive relearning. We demonstrate the effectiveness of our approach on the task of monocular and multi-view tracking and show that our approach outperforms the state-of-the-art.

Author Information

Angela Yao (National University of Singapore)
Juergen Gall (ETH Zurich)
Luc V Gool (Computer Vision Lab, ETH Zurich)
Raquel Urtasun (University of Toronto)

More from the Same Authors