A Survey on Human Action Recognition Using Depth Sensors

Liang, Bin; Zheng, Lihong

doi:10.1109/dicta.2015.7371223

Cited by 39 publications

(22 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…State-of-the-art methods for human action recognition using depth sensors are mostly based on depth maps, skeleton joints, or both [10]. Some of the advantages of depth sensors over color cameras are their invariance to lightning and color conditions.…”

Section: Related Workmentioning

confidence: 99%

Learning features combination for human action recognition from skeleton sequences

Luvizon

Tabia

Picard

2017

Pattern Recognition Letters

View full text Add to dashboard Cite

Human action recognition is a challenging task due to the complexity of human movements and to the variety among the same actions performed by distinct subjects. Recent technologies provide the skeletal representation of human body extracted in real time from depth maps, which is a high discriminant information for efficient action recognition. In this context, we present a new framework for human action recognition from skeleton sequences. We propose extracting sets of spatial and temporal local features from subgroups of joints, which are aggregated by a robust method based on the VLAD algorithm and a pool of clusters. Several feature vectors are then combined by a metric learning method inspired by the LMNN algorithm with the objective to improve the classification accuracy using the nonparametric k-NN classifier. We evaluated our method on three public datasets, including the MSR-Action3D, the UTKinect-Action3D, and the Florence 3D Actions dataset. As a result, the proposed framework performance overcomes the methods in the state of the art on all the experiments.

show abstract

Section: Related Workmentioning

confidence: 99%

Learning features combination for human action recognition from skeleton sequences

Luvizon

Tabia

Picard

2017

Pattern Recognition Letters

View full text Add to dashboard Cite

show abstract

“…A side benefit of the proposed method is that, in addition to coding, the global 3D scene representation is susceptible to be used in tasks such as action detection [10], scene recognition [8] or scene labeling [33].…”

Section: Resultsmentioning

confidence: 99%

3D hierarchical optimization for multi-view depth map coding

Maceira

Varas

Morros

et al. 2017

Multimed Tools Appl

View full text Add to dashboard Cite

Depth data has a widespread use since the popularity of high resolution 3D sensors. In multi-view sequences, depth information is used to supplement the color data of each view. This article proposes a joint encoding of multiple depth maps with a unique representation. Color and depth images of each view are segmented independently and combined in an optimal Rate-Distortion fashion. The resulting partitions are projected to a reference view where a coherent hierarchy for the multiple views is built. A Rate-Distortion optimization is applied to obtain the nal segmentation choosing nodes of the hierarchy. The consistent segmentation is used to robustly encode depth maps of multiple views obtaining competitive results with HEVC coding standards.Peer ReviewedPostprint (author's final draft

show abstract

“…The use of depth information has shown several advantages in a number of visual recognition tasks including human activity recognition [46]. Compared with RGB video sequences, depth information have shown several advantages in the context of activity recognition, for instance Liang and Zheng [47] claim that depth data can provide 3D structural information so that the motion information of activities can be more discriminative.…”

Section: Depth Information Estimationmentioning

confidence: 99%

Magnitude-Orientation Stream network and depth information applied to activity recognition

Caetano

Melo

Brémond

et al. 2019

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

The temporal component of videos provides an important clue for activity recognition, as a number of activities can be reliably recognized based on the motion information. In view of that, this work proposes a novel temporal stream for two-stream convolutional networks based on images computed from the optical flow magnitude and orientation, named Magnitude-Orientation Stream (MOS), to learn the motion in a better and richer manner. Our method applies simple non-linear transformations on the vertical and horizontal components of the optical flow to generate input images for the temporal stream. Moreover, we also employ depth information to use as a weighting scheme on the magnitude information to compensate the distance of the subjects performing the activity to the camera. Experimental results, carried on two well-known datasets (UCF101 and NTU), demonstrate that using our proposed temporal stream as input to existing neural network architectures can improve their performance for activity recognition. Results demonstrate that our temporal stream provides complementary information able to improve the classical two-stream methods, indicating the suitability of our approach to be used as a temporal video representation.

show abstract

A Survey on Human Action Recognition Using Depth Sensors

Cited by 39 publications

References 40 publications

Learning features combination for human action recognition from skeleton sequences

Learning features combination for human action recognition from skeleton sequences

3D hierarchical optimization for multi-view depth map coding

Magnitude-Orientation Stream network and depth information applied to activity recognition

Contact Info

Product

Resources

About