Strong Appearance and Expressive Spatial Models for Human Pose Estimation

By Leonid Pishchulin, Mykhaylo Andriluka, Peter Gehler, Bernt Schiele. 2013

222 citations recorded by the source corpus.

Typical approaches to articulated pose estimation combine spatial modelling of the human body with appearance modelling of body parts. This paper aims to push the state-of-the-art in articulated pose estimation in two ways. First we explore various types of appearance representations aiming to substantially improve the body part hypotheses. And second, we draw on and combine several recently proposed powerful ideas such as more flexible spatial models as well as image-conditioned spatial models. In a series of experiments we draw several important conclusions: (1) we show that the proposed appearance representations are complementary, (2) we demonstrate that even a basic tree-structure spatial human body model achieves state-of-the-art performance when augmented with the proper appearance representation, and (3) we show that the combination of the best performing appearance model with a flexible image-conditioned spatial model achieves the best result, significantly improving over the state of the art, on the ``Leeds Sports Poses'' and ``Parse'' benchmarks.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Convolutional Pose Machines (2016, background): “The classical approach to articulated pose estimation is the pictorial structures model [2, 3, 9, 14, 26, 27, 30, 43] in which spatial correlations between parts of the body are expressed as a tree-structured graphical model with kinematic priors that couple connected limbs.”
Stacked Hourglass Networks for Human Pose Estimation (2016, background): “Early work tackles such diﬃculties using robust image features and sophisticated structured prediction [1,2,3,4,5,6,7,8,9]: the former is used to produce local interpretations, whereas the latter is used to infer a globally consistent pose.”
Human Pose Estimation with Iterative Error Feedback (2015, background): “This is desirable in order to attack problems with higher-dimensional output spaces such as 3D human pose estimation [32, 33] or segmentation.”
Articulated Pose Estimation by a Graphical Model with Image Dependent Pairwise Relations (2014, positive): “Our method improves on all parts by a significant margin, and outperforms the best previously published result [16] by 5.”
Towards a Visual Turing Challenge (2014, background): “Recently we witness a tremendous progress in the machine per c ption [1, 2, 3, 4, 5, 6, 7, 8] and in the language understanding [9, 10, 11, 12, 13] tasks.”
Recurrent Human Pose Estimation (2016, neutral): “ed) 94.3 87.1 78.6 72.0 78.1 83.2 77.1 81.5 Ours (MPII & LSP, 1 it.) 95.6 88.8 80.7 75.5 83.0 86.2 80.6 84.3 Ours (MPII & LSP, 2 it.) 95.2 89.0 81.5 77.0 83.7 87.0 82.8 85.2 Pishchulin et al. [28] 87.2 56.7 46.7 38.0 61.0 57.5 52.7 57.1 Wang Li et al. [42] 84.7 57.1 43.7 36.7 56.7 52.4 50.8 ”
Pose Machines: Articulated Pose Estimation via Inference Machines (2014, methodology): “In [20], the authors use a strong appearance model, by training rotation dependent part detectors with separate part detectors for the head and torso while using a simple tree-structured model.”
Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation (2018, unknown): “[27] 74.”
Towards Viewpoint Invariant 3D Human Pose Estimation (2016, background): “In contrast, discriminative methods directly identify individual body parts, labels, or positions and construct the skeleton in a bottom-up approach [51,52,14,54,15].”
VNect (2017, background): “Monocular RGB body pose estimation in 2D has been widely researched, but”
Fine-Grained Activity Recognition with Holistic and Pose Based Features (2014, background): “However, articulated pose estimation is a challenging and non-trivial task that is subject of ongoing research [18,19,32,8,24].”
Multi-source Deep Learning for Human Pose Estimation (2014, background): “On the other hand, many recent works use local body parts [52, 54, 11, 58, 9, 13, 48, 40, 2, 42, 21, 46, 55, 41, 1] in order to handle the many degrees of freedom in body part articulation.”
Attribute And-Or Grammar for Joint Parsing of Human Attributes, Part and Pose (2016, background): “In the last few years, hierarchical mixture models [6], [29], [37], [38] made significant progress which is similar to a dependency grammar.”
Pose for Action - Action for Pose (2016, background): “Pose estimation in videos can be improved by taking temporal information or motion cues into account [2], [10], [12], [14], [16], [45].”