Real-time classification of dance gestures from skeleton animation

By Michalis Raptis, Darko Kirovski, Hugues Hoppe. Symposium on Computer Animation, 2011

341 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Control and Systems Engineering, Signal Processing

View Paper: https://www.semanticscholar.org/paper/bc404e002441e7aaa445f6ef62ab8d9420ff9ffc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

In Signal Processing

Citations in the corpus

Learning Actionlet Ensemble for 3D Human Action Recognition (2014, neutral): “[27] utilize Kinect cameras to recognizes dance actions.”
Classifying and visualizing motion capture sequences using deep neural networks (2013, background): “Conventional designs may be classiﬁed into two categories: a whole motion is represented by one feature matrix or vector [14], and classiﬁed by a classiﬁer as a whole [10]; or a library of key features [18] is learned from the whole dataset, and then each motion is represented as a bag or histogram…”
Online RGB-D gesture recognition with extreme learning machines (2013, background): “In [23], dance gestures are classified from Kinect skeletal data.”
Spatiotemporal representation of 3D skeleton joints-based action recognition using modified spherical harmonics (2016, methodology): “The calculation of our features are computationally light than many previous works which try to classify a large 3D point, such in (Raptis et al., 2011), (Raviteja et al.)”
Human gesture recognition using Kinect camera (2012, methodology): “A multicategory SVM is used to classify human postures using a human star skeleton, angles of six sticks in the star skeleton, and object motion vectors.”
Gesture Recognition Portfolios for Personalization (2014, background): “A number of gesture recognition methods have been proposed in the literature [15, 6, 13], including some which operate on the human pose estimated from the Kinect [15, 6].”
3D skeleton-based human action classification: A survey (2016, methodology): “In particular, a descriptor for the skeletal data similar to that proposed in [118] is adopted in order to obtain a representation robust to orientation changes.”
Body Surface Context: A New Robust Feature for Action Recognition From Depth Videos (2014, methodology): “[30] used an angular representation of the skeleton for dance gesture classification.”
Real-Time Gesture Recognition from Depth Data through Key Poses Learning and Decision Forests (2012, methodology): “Our representation is an improvement of the work of Raptis et al.[3].”
Action recognition from motion capture data using Meta-Cognitive RBF Network classifier (2014, background): “Not untill recently numerous researchers have started using MoCap data for action recognition [1] and [2].”
Exploring the Trade-off Between Accuracy and Observational Latency in Action Recognition (2012, methodology): “Traditionally, accuracy has driven the design of recognition systems.”
Effective Active Skeleton Representation for Low Latency Human Action Recognition (2016, background): “[17], [18] can only handle segmented videos, i.”
The Moving Pose: An Efficient 3D Kinematics Descriptor for Low-Latency Action Recognition and Detection (2013, methodology): “The accurate real-time tracking of 3D skeletons [21], made possible with the introduction of low-cost RGB-D cameras, led to the development of efficient methods for classification of dance moves [17] and other arbitrary actions [7, 24].”
Ongoing human action recognition with motion capture (2014, background): “[30] have used joint angles as features to recognize dance actions in a game-based application.”
Validity of the Microsoft Kinect for assessment of postural control. (2012, background): “The results of previous studies are promising, and have shown that the depth sensor itself is accurate for assessing 3D position in a workplace environment [14], and that joint centers derived from the Microsoft Kinect can be used to classify dance gestures [15].”
Pose-based human action recognition via sparse representation in dissimilarity space (2014, background): “Another transformat ion of joint coordinates which was recently proposed in [26] is the ‘‘ Torso PCA frame’’ (TPCAF), which is the representat ion of choice in the current work.”