Using dynamic programming to match human behavior sequences

Chen, Yan; Wu, Qiang; He, Xiangjian

doi:10.1109/icarcv.2008.4795746

Cited by 6 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, because of the endpoint localization issue, DP cannot be applied directly to recognize gestures in a continuous stream. Dynamic Time Warping (DTW), which is an application of DP and has been widely used in gesture recognition, has unfortunately only been applied to isolated gesture recognition [27,28,12].…”

Section: Continuous Gesture Recognition Using Dynamic Programmingmentioning

confidence: 99%

“…While the first method combines a motion detection technique and an explicit multi-scale search to find the start and end times of a gesture, the other two methods further improve the efficiency and accuracy of Method I by employing a Dynamic Programming (DP) [25] technique. In particular, the second algorithm extends Dynamic Time Warping (DTW) [26], which is a special application of DP and can only be applied to isolated gesture recognition due to its endpoint constraint [27,28,12], to the case of continuous gesture recognition. Note that Method II is very different from the traditional LevelBuilding DTW, which was initially proposed by Myers and Rabiner [29] for connected word recognition, and was recently enhanced for continuous sign recognition [30,31].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Model-based segmentation and recognition of dynamic gestures in continuous video streams

Greenspan

2011

Pattern Recognition

View full text Add to dashboard Cite

Section: Continuous Gesture Recognition Using Dynamic Programmingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Model-based segmentation and recognition of dynamic gestures in continuous video streams

Greenspan

2011

Pattern Recognition

View full text Add to dashboard Cite

“…To align actions that are linear sequence of poses, which may vary in temporal execution, sequence alignment techniques such as Dynamic Time Warping (DTW) [18,23], Dynamic Manifold Warping (DMW) [14,15], and Canonical Time Warping (CTW) [29], have been employed to reduce the impact of temporal variations, [5,6,10,22]. These methods have however been criticized in situations where the temporal execution rate may provide some key information between two classes, e.g.…”

Section: Introductionmentioning

confidence: 99%

Generating Local Temporal Poses from Gestures with Aligned Cluster Analysis for Human Action Recognition

Edwards¹,

Xie²

2015

Procedings of the Proceedings of the 7th UK British Machine Vision Workshop 2015

View full text Add to dashboard Cite

The use of pose estimation for human action recognition has seen a resurgence in previous years, due in part to the natural representation of the activity as a sequence of key poses and gestures. The use of sequence alignment techniques has aided the process of comparing between sequences of differing temporal rates, with aligned cluster analysis segmenting an observation into lower level action primitives. We suggest that the representation of a given action class via its lower level gestures can help to identify the higher-level action class label. We therefore present a method for the generation of key poses via the initial segmentation of an action class into gestures that are similar across numerous observations. We treat all training observations as a single observation in which there are repetitions of the same action class. By applying segmentation, we then identify common gestures across the class, which are used to generate the key poses we optimize via evolutionary programming. Global recognition rates of 97.4% are achieved using a subset of the MSR Action3D dataset. We then expand the method to recognize interaction events between two individuals using the SBU Kinect Interaction dataset, achieving recognition rates of 83.9% and over 96.4% when observing the first 6 classes.

show abstract