PoseField: An Efficient Mean-Field Based Method for Joint Estimation of Human Pose, Segmentation, and Depth

Abstract

Abstract. Many models have been proposed to estimate human pose and segmentation by leveraging information from several sources. A standard approach is to formulate it in a dual decomposition framework. However, these models generally suffer from the problem of high computational complexity. In this work, we propose PoseField, a new highly efficient filter-based mean-field inference approach for jointly estimating human segmentation, pose, per-pixel body parts, and depth given stereo pairs of images. We extensively evaluate the efficiency and accuracy offered by our approach on H2View [1], and Buffy [2] datasets. We achieve 20 to 70 times speedup compared to the current state-of-the-art methods, as well as achieving better accuracy in all these cases.

Deep Learning Markov Random Field for Semantic Segmentation

“…Markov Random Field (MRF) or Conditional Random Field (CRF) has achieved great successes in semantic image segmentation, which is one of the most challenging problems in computer vision. Researchers improved labeling accuracy by exploring rich information to define the pairwise functions, including long-range dependencies [18], [37], high-order potentials [19], [38], and semantic label contexts [3], [39], [20]. For example, Krähenbühl et al [18] attained accurate segmentation boundary by inferring on a fully-connected graph.…”

Semantic Instance Annotation of Street Scenes by 3D to 2D Label Transfer

“…This mean field approximation can be computed efficiently using bilateral filtering [22]. As our model comprises three sets of densely connected variables (namely P, L and P ↔ L), we exploit the algorithm of [21,45] which generalizes [22] to multiple fields. Learning: We employ empirical risk minimization in order to learn the parameters in our model, considering the univariate logistic loss, defined as ∆(s) = − log (P (s)) where P (·) denotes the marginal distribution at the respective site.…”

Efficient 2D human pose estimation using mean-shift

“…We compared our extended model with four recent 2D pose extraction algorithms [2,12,16] and [15] on Buffy dataset using the values reported in [15] (see Table 2). …”

“…Eichner et al [2] and Vineet et al [15] propose a search space reduction methodology by utilizing image segmentation techniques to extract foreground objects. However, this has deteriorating effect on the edge features commonly used for limb detection.…”