Latent Pose Estimator for Continuous Action Recognition

Ning, Hui; Xu, Wei; Gong, Yi; Huang, Thomas S.

doi:10.1007/978-3-540-88688-4_31

Cited by 37 publications

(34 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, a GMM is used to globally model the postures, and an action graph [11] is used for inference. On the other hand, parallel to the approaches developed for temporal modelling of human actions in color videos such as [15,2], Lv and Nevatia in [14] employ a hidden Markov model (HMM) to represent the transition probability for pre-defined 3D joint positions. Similarly, in [8], the 3D joint position is described using another generative model, which is a conditional random field (CRF).…”

Section: Related Workmentioning

confidence: 99%

HON4D: Histogram of Oriented 4D Normals for Activity Recognition from Depth Sequences

Oreifej

Liu

2013

2013 IEEE Conference on Computer Vision and Pattern Recognition

854

868

View full text Add to dashboard Cite

We present a new descriptor for activity recognition from videos acquired by a depth sensor. Previous descriptors mostly compute shape and motion features independently; thus, they often fail to capture the complex joint shapemotion cues at pixel-level. In contrast, we describe the depth sequence using a histogram capturing the distribution of the surface normal orientation in the 4D space of time, depth, and spatial coordinates. To build the histogram, we create 4D projectors, which quantize the 4D space and represent the possible directions for the 4D normal. We initialize the projectors using the vertices of a regular polychoron. Consequently, we refine the projectors using a discriminative density measure, such that additional projectors are induced in the directions where the 4D normals are more dense and discriminative. Through extensive experiments, we demonstrate that our descriptor better captures the joint shape-motion cues in the depth sequence, and thus outperforms the state-of-the-art on all relevant benchmarks.

show abstract

Section: Related Workmentioning

confidence: 99%

HON4D: Histogram of Oriented 4D Normals for Activity Recognition from Depth Sequences

Oreifej

Liu

2013

2013 IEEE Conference on Computer Vision and Pattern Recognition

854

868

View full text Add to dashboard Cite

show abstract

“…According to [19], estimating human poses from 2D video is harsh due to large variations in appearance. In addition, the segmentation of human figures in order to estimate the pose in RGB images is very computationally expensive, due to the high dimensionality of visual features [20].…”

Section: Har In Rgb Videomentioning

confidence: 99%

Flexible human action recognition in depth video sequences using masked joint trajectories

Tejero-de-Pablos

Nakashima

Yokoya

et al. 2016

J Image Video Proc.

View full text Add to dashboard Cite

Human action recognition applications are greatly benefited from the use of commodity depth sensors that are capable of skeleton tracking. Some of these applications (e.g., customizable gesture interfaces) require learning of new actions at runtime and may not count with many training instances. This paper presents a human action recognition method designed for flexibility, which allows taking users' feedback to improve recognition performance and to add a new action instance without computationally expensive optimization for training classifiers. Our nearest neighbor-based action classifier adopts dynamic time warping to handle variability in execution rate. In addition, it uses the confidence values associated to each tracked joint position to mask erroneous trajectories for robustness against noise. We evaluate the proposed method with various datasets with different frame rates, actors, and noise. The experimental results demonstrate its adequacy for learning of actions from depth sequences at runtime. We achieve an accuracy comparable to the state-of-the-art techniques on the challenging MSR-Action3D dataset.

show abstract

“…It is widely believed that 3D pose estimation is sufficiently noisy that estimator bias and variance will outweigh the benefits of such compelling representations. Nevertheless, some recent methods have successfully demonstrated that this may not be the case (e.g., [22]). Unlike such work focused on classifying very different motion patterns, we tackle the more subtle problem of inferring meaningful percepts from locomotion.…”

Section: Background and Related Workmentioning

confidence: 99%