Online human gesture recognition from motion data streams

By Xin Zhao, Xue Li, Chaoyi Pang, Xiaofeng Zhu, Quan Z. Sheng. 2013

91 citations recorded by the source corpus.

Subfields: Biomedical Engineering, 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 Biomedical Engineering

In Computer Vision and Pattern Recognition

In Human-Computer Interaction

Citations in the corpus

Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, methodology): “Bag-of-words encoding is widely employed by a large number of skeleton-based human representations [100], [106], [118], [134], [135], [138], [141], [143], [149], [152], [155], [157], [165], [170], [177], [179],”
Adaptive RNN Tree for Large-Scale Human Action Recognition (2017, background): “[30] extracted structured streaming skeleton features to represent actions and constructed classifiers based on sparse coding features.”
A discussion on the validation tests employed to compare human action recognition methods using the MSR Action3D dataset (2014, unknown): “[41] 2013 - 81.”
Effective Active Skeleton Representation for Low Latency Human Action Recognition (2016, methodology): “These states can be obtained by a clustering method [13], [14] (e.”
Leveraging Hierarchical Parametric Networks for Skeletal Joints Based Action Segmentation and Recognition (2014, unknown): “576 Structured Streaming Skeletons [32] 0.”
Online Action Recognition via Nonparametric Incremental Learning (2014, methodology): “As it handles videos on a frame-by-frame basis, the method is suitable to tackle the so-called “continuous action recognition” problem [3, 4, 18, 36, 46, 52], in which both training and test videos contain a number of gestures in a sequence.”
Clustered Spatio-temporal Manifolds for Online Action Recognition (2014, methodology): “[5] optimise the size of the segment during their feature extraction using a DTW variant for subsequence matching.”
Encouraging LSTMs to Anticipate Actions Very Early (2017, methodology): “This was followed by several attempts to model the dynamics of the observed actions, such as by introducing integral and dynamic bag-of-words [25], using spatial-temporal implicit shape models [45], extracting human body movements via skeleton information [47], and accounting for the complete and pa”
Temporal Order-Preserving Dynamic Quantization for Human Action Recognition from Multimodal Sensor Streams (2015, methodology): “Many temporal modeling methods [7, 20, 14] suffer from the temporal misalignment problem due to the variations in the execution rate.”
Spectral Graph Skeletons for 3D Action Recognition (2014, methodology): “DL-GSGC [18] 96.7 MMTW [30] 92.7 MP [35] 91.7 SGS ( p mean ) 91.4 HOD [11] 90.2 HON4D [21] 88.9 AE [29] 88.2 SGS ( p max ) 86.3 SSS [36] 81.7 Canonical poses [10] 65 82.2 DMM-HOG [34] 66.1 98.8 MP [35] 98.5 Canonical poses [10] 95.9 to sparse coding, as previously noted in Sec.”
Predicting the Where and What of Actors and Actions through Online Action Localization (2016, background): “[45] represent the structure of streaming skeletons (poses) by a combination of humanbody-part movements and use it to recognize actions in RGB-D.”
Category-Blind Human Action Recognition: A Practical Recognition System (2015, methodology): “As in [11, 22, 25], the optimal θε is chosen such that it minimizes the recognition error, e.”
Real-Time Multi-scale Action Detection from 3D Skeleton Data (2015, methodology): “However, both [26, 27] require providing a label for every frame during training and testing.”