Parsing occluded people by flexible compositions

By Xianjie Chen, Alan L. Yuille. 2014

81 citations recorded by the source corpus.

This paper presents an approach to parsing humans when there is significant occlusion. We model humans using a graphical model which has a tree structure building on recent work [32, 6] and exploit the connectivity prior that, even in presence of occlusion, the visible nodes form a connected subtree of the graphical model. We call each connected subtree a flexible composition of object parts. This involves a novel method for learning occlusion cues. During inference we need to search over a mixture of different flexible models. By exploiting part sharing, we show that this inference can be done extremely efficiently requiring only twice as many computations as searching for the entire object (i.e., not modeling occlusion). We evaluate our model on the standard benchmarked “We Are Family” Stickmen dataset and obtain significant performance improvements over the best alternative algorithms.

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

Recent advances in convolutional neural networks (2015, methodology): “[151], [152] also combine graphical model with CNN.”
DeeperCut: A Deeper, Stronger, and Faster Multi-person Pose Estimation Model (2016, methodology): “For any pair of detections (d, d′) (Fig.”
DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimation (2015, methodology): “Additional Results on WAF dataset Qualitative comparison of our joint formulation DeepCut MP Dense-CNN to the traditional two-stage approach Dense-CNN det ROI relying on person detector, and to the approach of Chen&Yuille [2] on WAF dataset is shown in Fig.”
RMPE: Regional Multi-person Pose Estimation (2016, methodology): “Head Shoulder Elbow Wrist Total Chen&Yuille [4] 83.”
Multi-person Pose Estimation with Local Joint-to-Person Associations (2016, background): “For every detected person, [13] explores a range of tree structured models each containing only a subset of upper-body parts, and selects the best model based on a cost function that penalizes a model containing occluded parts.”
DeepVoting: An Explainable Framework for Semantic Part Detection under Partial Occlusion (2017, methodology): “In this scenario, either a ﬂexible model is applied to allow the absence of some components (Chen et al. 2014), or the hidden parts need to be inferred based on the spatial or geometric cues of other parts (Chen and Yuille 2015).”
Visual Concepts and Compositional Voting (2017, background): “As has already been shown, part-based models are very useful for detecting partially occluded objects [7, 6, 21].”
CrowdPose: Efficient Crowded Scenes Pose Estimation and a New Benchmark (2018, methodology): “Globally Optimizing Association Next, we compare our association algorithm to several NMS algorithms, including bounding box NMS, poseNMS [3, 7] and parametric poseNMS [13].”
Deep Learning-based Human Pose Estimation: A Survey (2020, methodology): “intuitive human body model is successfully utilized in 2D HPE [16] [17] and 3D HPE [18] [19].”
Deep Occlusion-Aware Instance Segmentation with Overlapping BiLayers (2021, unknown): “Occlusion Handling Methods for occlusion handling have been proposed [50, 57, 17, 10, 60, 23, 17, 65, 59].”
AlphaPose: Whole-Body Regional Multi-Person Pose Estimation and Tracking in Real-Time (2022, background): “For the data transform module, we implement vanilla box NMS and soft-NMS [83].”