Discriminative human action recognition in the learned hierarchical manifold space

By Lei Han, Xinxiao Wu, Wei Liang, Guangming Hou, Yunde Jia. 2010

103 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition, Control and Systems Engineering

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

1. Yingli Tian
2. Fei Han
3. Xiaodong Yang
4. Hao Zhang
5. William A. Hoff
6. Brian Reily
7. Guo-Jun Qi
8. Jun Liu

In Computer Vision and Pattern Recognition

1. Yingli Tian
2. Luc Van Gool
3. Wanqing Li
4. Pichao Wang
5. G. Wang
6. Ming-Hsuan Yang
7. Jun Liu
8. Dacheng Tao

In Control and Systems Engineering

1. William A. Hoff
2. Brian Reily
3. Fei Han
4. Hao Zhang
5. Alois C. Knoll
6. Guang Chen
7. N. Nikolaidis
8. Juergen Gall

Citations in the corpus

• Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, methodology): “[182] Hierarchical Manifolds BoW Manif Dict X X X X Slama et al.”
• Effective 3D action recognition using EigenJoints (2014, methodology): “The MoCap system [17] was used to extract 3D joint positions by using markers and high precision camera array.”
• A Survey on Human Motion Analysis from Depth Data (2013, background): “(b) Example of a typical hierarchy of human body parts in a tree structure as in [22] c ©2010 Elsevier .”
• Action recognition on motion capture data using a dynemes and forward differences representation (2014, methodology): “A hierarchical discriminative approach is used in [18] by Han et al.”
• HON4D: Histogram of Oriented 4D Normals for Activity Recognition from Depth Sequences (2013, methodology): “Similarly, in [8], the 3D joint position is described using another generative model, which is a conditional random field (CRF).”
• Temporal Order-Preserving Dynamic Quantization for Human Action Recognition from Multimodal Sensor Streams (2015, methodology): “Other methods such as graph model-based methods [6, 2] and motion template-based methods [7, 20, 10] emphasize on the modeling of the temporal dynamics.”
• Capturing Global and Local Dynamics for Human Action Recognition (2014, background): “Conditional random fields [7] and hidden Markov models [12] are common examples of statistical models.”
• Enhanced Sequence Matching for Action Recognition from 3D Skeletal Data (2014, unknown): “Hidden Markov Model (HMM) [16–18, 8] and Conditional Random Field (CRF) [19–21] are popular models for this approach.”
• Ongoing human action recognition with motion capture (2014, background): “[13] have exploited the skeleton hierarchy to compute trajectories, where actions were represented in a manifold space.”
• Learning Actionlet Ensemble for 3D Human Action Recognition (2014, unknown): “[13] used CRF over 3D joint positions.”
• Classifying actions based on histogram of oriented velocity vectors (2015, background): “…(Huang et al. 2010; Kilner et al. 2009; Pehlivan and Duygulu 2011; Song et al. 2011; Veeraraghavan et al. 2009); however, those systems are limited to structured environments (Han et al. 2010; Oi et al. 2013) and they would be impractical for some applications such as daily human-robot interaction.”
• Range-Sample Depth Feature for Action Recognition (2014, methodology): “In [3], conditional random field (CRF) was used for action recognition.”
• Action recognition using ensemble weighted multi-instance learning (2014, background): “An ISA network [6] is described as a two-layer neural network, with square and square-root nonlinearities in the first and second layers respectively.”
• Joint Angles Similarities and HOG2 for Action Recognition (2013, unknown): “Alternatively, other common motion models may utilize a hidden Markov model (HMM) [11], or a conditional random field (CRF) [5].”