Human Pose Estimation Using Exemplars and Part Based Refinement

Abstract

Abstract. In this paper, we proposed a fast and accurate human pose estimation framework that combines top-down and bottom-up methods. The framework consists of an initialization stage and an iterative searching stage. In the initialization stage, example based method is used to find several initial poses which are used as searching seeds of the next stage. In the iterative searching stage, a larger number of body parts candidates are generated by adding random disturbance to searching seeds. Belief Propagation (BP) algorithm is applied to these candidates to find the best n poses using the information of global graph model and part image likelihood. Then these poses are further used as searching seeds for the next iteration. To model image likelihoods of parts we designed rotation invariant EdgeField features based on which we learnt boosted classifiers to calculate the image likelihoods. Experiment result shows that our framework is both fast and accurate.

Who are like me: Fast human pose retrieval in unconstrained environments

“…The similarity is given by the sum of fraction of the distance between corresponding endpoints and the length of the corresponding part. We use pose estimation algorithm [14], which adopts strong priors to initialize potential body part locations and is fast with competitive accuracy similar to [6]. The reason why we select [14] is that it uses strong priors from the annotated pose to initialize body parts which do benefit in retrieval rate and speed comparing with other pose estimation algorithms.…”

“…We use pose estimation algorithm [14], which adopts strong priors to initialize potential body part locations and is fast with competitive accuracy similar to [6]. The reason why we select [14] is that it uses strong priors from the annotated pose to initialize body parts which do benefit in retrieval rate and speed comparing with other pose estimation algorithms. Note that as there is no information of human positions, we manually crop people out of the natural environments from each image.…”