Real-time human pose recognition in parts from single depth images

By Jamie Shotton, Toby Sharp, Alex Kipman, Andrew William Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, Richard Moore. 2011

5,429 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Human-Computer Interaction

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

In Human-Computer Interaction

Citations in the corpus

Robust Neural Machine Translation with Joint Textual and Phonetic Embedding (2018, background): “Data augmentation is widely known to be extremely useful in training models (Shotton et al., 2011; Krizhevsky et al., 2012), especially in training data-hungry neural networks.”
Conditional Random Fields as Recurrent Neural Networks (2015, methodology): “We apply the proposed method to the problem of semantic image segmentation, obtaining top results on the challenging Pascal VOC 2012 segmentation benchmark.”
Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation (2014, methodology): “A technique that works for single depth images, Shotton et al.’s body part detector [32] uses a random forest of decision trees to create simple body part detectors followed by skeletal processing.”
Learning Rich Features from RGB-D Images for Object Detection and Segmentation (2014, methodology): “[31], which generalize those originally proposed by Geman et al.”
Stacked Hourglass Networks for Human Pose Estimation (2016, background): “There are variations to the pose estimation problem which include use of additional features such as depth or motion cues [27,28].”
DeepCoder: Learning to Write Programs (2016, background): “4.5 provides theoretical indication that ignoring the correlations between functions is not cataclysmic for Sort and add enumeration, since a Rank loss that upper bounds the Sort and add runtime can still be minimized.”
NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity Understanding (2019, background): “To have view-invariant representations of the actions, features can be extracted from the 3D body joint positions [51], [52] which are available for each frame.”
VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (2017, methodology): “Some of the earlier methods for feature representation are [41, 8, 7, 19, 42, 35, 6, 27, 1, 36, 2, 25, 26].”
Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, methodology): “[20] in 2011 provided an extremely effective skeleton construction algorithm based on body part recognition, that was able to work in real time.”
Multimodal Multipart Learning for Action Recognition in Depth Videos (2015, background): “This is done by a body part detection on depth sequences of human body movements [1].”
Histogram of Oriented Principal Components for Cross-View Action Recognition (2014, positive): “In [37], we showed how HOPC outperforms existing techniques for front view action recognition.”
HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition (2014, methodology): “On the other hand, some methods [36,31,37] use the human joint positions extracted by the OpenNI tracking framework (OpenNI) [24] as interest points.”
Cross-View Action Modeling, Learning, and Recognition (2014, background): “For a human subject, 21 joint positions are tracked by the skeleton tracker [17] and each joint i has 3 location coordinates pj(t) = (xj(t), yj(t), zj(t)), a motion vector mj(t) = (∆xj(t),∆yj(t),∆zj(t)) as well the visibility label hj(t) at a frame t.”
Learning human activities and object affordances from RGB-D videos (2012, background): “Building upon the recent advances in human pose detection from an RGB-D sensor (Shotton et al., 2011), in this paper we present learning algorithms to detect the human activities and object affordances.”
Feature Boosting Network For 3D Pose Estimation (2019, background): “Most of the existing works make use of highly constrained configurations [2], such as multi-view systems [3] and depth sensors [4], to infer the 3D poses.”
RGB-D-based Human Motion Recognition with Deep Learning: A Survey (2017, background): “Additionally, 3D positions of body joints can be estimated from depth maps [114].”
Modeling Temporal Dynamics and Spatial Configurations of Actions Using Two-Stream Recurrent Neural Networks (2017, methodology): “[38] propose a method to quickly and accurately predict 3D positions of body joints from a single depth image.”
View Adaptive Recurrent Neural Networks for High Performance Human Action Recognition from Skeleton Data (2017, background): “Besides, the prevalence of cost-effective depth cameras such as Microsoft Kinect [48], Intel RealSense [1], dual camera devices, and the advance of a powerful technique of human pose estimation from depth [34] make 3D skeleton data easily obtainable.”