Grammar of Dance Gesture from Bali Traditional Dance
01 Jan 2012-
TL;DR: The empiric results showed that probabilistic grammar-based classifiers that were induced using the Alergia algorithm with Symbolic Aggregate Approximation (SAX) discretization method achieved 92% of average precision in recognizing a predefined set of dance gestures.
Abstract: Automatic recognition of dance gesture is one important research area in computer vision with many potential applications. Bali traditional dance comprises of many dance gestures that relatively unchanged over the years. Although previous studies have reported various methods for recognizing gesture, to the best of our knowledge, a method to model and classify dance gesture of Bali traditional dance is still unfound in literature. The aim of this paper is to build a robust recognizer based on linguistic motivated method to recognize dance gesture of Bali traditional dance choreography. The empiric results showed that probabilistic grammar-based classifiers that were induced using the Alergia algorithm with Symbolic Aggregate Approximation (SAX) discretization method achieved 92% of average precision in recognizing a predefined set of dance gestures. The study also showed that the most discriminative features to represent Bali traditional dance gestures are skeleton joint features of: left/right
Citations
More filters
TL;DR: This survey is to present a comprehensive survey on automated dance gesture recognition with emphasis on static hand gesture recognition, considering human hands because human hands are the most flexible part of the body and can transfer the most meaning.
Abstract: Gesture recognition means the identification of different expressions of human body parts to express the idea, thoughts and emotion. It is a multi-disciplinary research area. The application areas of gesture recognition have been spreading very rapidly in our real-life activities including dance gesture recognition. Dance gesture recognition means the recognition of meaningful expression from the different dance poses. Today, research on dance gesture recognition receives more and more attention throughout the world. The automated recognition of dance gestures has many applications. The motive behind this survey is to present a comprehensive survey on automated dance gesture recognition with emphasis on static hand gesture recognition. Instead of whole body movement, we consider human hands because human hands are the most flexible part of the body and can transfer the most meaning. A list of research issues and open challenges is also highlighted.
17 citations
Cites background or methods from "Grammar of Dance Gesture from Bali ..."
...In [9], emphasis has been given on Bali traditional dance....
[...]
...Their work is carried with only eleven co-ordinates out of twenty different joints of skeleton,(iii)Bali Traditional Dance [9], which works on probabilistic grammar-based classifier, (iv) Ballet Dance [22] where multiple stage system is proposed to recognize the different dance posture, (v) Kazakh Traditional Dance [26] is basically concerned with the head gestures and many others....
[...]
...Kinect sensor camera [19] or depth sensor camera [9] are used for capturing 3-D images....
[...]
...Also, it includes applications such as dance selfassessment and e-learning of dances [9], [10]....
[...]
02 Apr 2015
TL;DR: Fusion skeletal data obtained from two views for recognizing human action is used and Experimental results show that fusion skeletal data consistently give better recognition performance than their single view counterpart.
Abstract: Advancement of RGB-D cameras that are capable of tracking human body movement in the form of a skeleton has contributed to growing interest in skeleton-based human action recognition. However, the tracking performance of a single camera is prone to occlusion and is view dependent. In this study, we use fusion skeletal data obtained from two views for recognizing human action. We perform a substitutive fusion based on joint tracking status and build a view-invariant action recognition system. The resulting fusion skeletal data are transformed into histogram of cubes as a frame level feature. Clustering is applied to build a dictionary of frame representatives, and actions are encoded as sequences of frame representatives. Finally, recognition is performed as a sequence matching task by using Dynamic Time Warping with K-nearest neighbor. Experimental results show that fusion skeletal data consistently give better recognition performance than their single view counterpart.
15 citations
Cites background from "Grammar of Dance Gesture from Bali ..."
...It has a wide range of real-world applications, including surveillance systems [1], human-computer interactions [2], ambient asisted living [3], and even culture [4], [5]....
[...]
01 Dec 2016
TL;DR: A simple two-level classification method for asamyukta hastas of Sattriya dance which is an Indian classical dance form is introduced and one of the applications can be in the e-learning and self learning of the dance hand gestures (mudras or hastas).
Abstract: The single-hand gestures of Indian classical dance are termed as ‘Asamyukta Hastas’ which is a combination of two Sanskrit words, asamyukta meaning ‘single’ and hastas meaning ‘hand gestures’. This paper introduces a simple two-level classification method for asamyukta hastas of Sattriya dance which is an Indian classical dance form. In the first level, twenty nine classes of hastas are categorized into three groups based on their structural similarity. Then, in the next level hastas are individually recognized from the database within the group. Moreover, the proposed method extracts Medial Axis Transformation (MAT) from the captured images to identify the groups in the first level. One of the applications of the outcome of this research work can be in the e-learning and self learning of the dance hand gestures (mudras or hastas).
12 citations
19 Apr 2016
TL;DR: An interface to visualize and analyze body movement data is proposed that enables users to navigate and explore the evolution of movement over time for different movement areas and a clustering method based on hierarchical clustering to group similar movement patterns is proposed.
Abstract: The advancement of motion sensing input devices has enabled the collection of multivariate time-series body movement data. Analyzing such type of data is challenging due to the large amount of data and the task of mining for interesting temporal movement patterns. To address this problem, we propose an interface to visualize and analyze body movement data. This visualization enables users to navigate and explore the evolution of movement over time for different movement areas. We also propose a clustering method based on hierarchical clustering to group similar movement patterns. The proposed visualization is illustrated with a case study which demonstrates the ability of the interface to analyze body movements.
10 citations
Cites background from "Grammar of Dance Gesture from Bali ..."
...With the (ii) approach, s1 and s2 are similar, because the corresponding event types are proportional ([10,20] is proportional to [20,40], [20,5] is proportional to [40,10], and [6,18] is proportional to [10,30])....
[...]
01 Aug 2016
TL;DR: This paper explores the problem of folk dances recognition from video recordings, focusing on Greek folk dances, using different representations for the data, including a representation based on Linear Dynamic Systems and a novel variant that uses dense trajectories descriptors instead of pixel intensities.
Abstract: Dance traditions constitute a significant aspect of cultural heritage around the world. The organization, semantic analysis, and retrieval of dance-related multimedia content (i.e., music, video) in databases is, therefore, crucial to their preservation. In this paper we explore the problem of folk dances recognition from video recordings, focusing on Greek folk dances, using different representations for the data. To this end we have employed the well-known Bag of Words model, in combination with dense trajectories, as well as with streaklines descriptors. Furthermore, we have adopted a representation based on Linear Dynamic Systems, including a novel variant that uses dense trajectories descriptors instead of pixel intensities. The performance of the aforementioned representations is evaluated and compared, in a classification scenario involving 13 different dance classes.
2 citations
Cites background from "Grammar of Dance Gesture from Bali ..."
...Certain methods are also concerned with the recognition of folk dances, which can be regarded as more relevant to our problem....
[...]
References
More filters
TL;DR: In this paper, a face detection framework that is capable of processing images extremely rapidly while achieving high detection rates is described. But the detection performance is limited to 15 frames per second.
Abstract: This paper describes a face detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the “Integral Image” which allows the features used by our detector to be computed very quickly. The second is a simple and efficient classifier which is built using the AdaBoost learning algorithm (Freund and Schapire, 1995) to select a small number of critical visual features from a very large set of potential features. The third contribution is a method for combining classifiers in a “cascade” which allows background regions of the image to be quickly discarded while spending more computation on promising face-like regions. A set of experiments in the domain of face detection is presented. The system yields face detection performance comparable to the best previous systems (Sung and Poggio, 1998; Rowley et al., 1998; Schneiderman and Kanade, 2000; Roth et al., 2000). Implemented on a conventional desktop, face detection proceeds at 15 frames per second.
13,037 citations
07 Jul 2001
TL;DR: A new image representation called the “Integral Image” is introduced which allows the features used by the detector to be computed very quickly and a method for combining classifiers in a “cascade” which allows background regions of the image to be quickly discarded while spending more computation on promising face-like regions.
Abstract: This paper describes a face detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our detector to be computed very quickly. The second is a simple and efficient classifier which is built using the AdaBoost learning algo- rithm (Freund and Schapire, 1995) to select a small number of critical visual features from a very large set of potential features. The third contribution is a method for combining classifiers in a "cascade" which allows back- ground regions of the image to be quickly discarded while spending more computation on promising face-like regions. A set of experiments in the domain of face detection is presented. The system yields face detection perfor- mance comparable to the best previous systems (Sung and Poggio, 1998; Rowley et al., 1998; Schneiderman and Kanade, 2000; Roth et al., 2000). Implemented on a conventional desktop, face detection proceeds at 15 frames per second.
10,592 citations
"Grammar of Dance Gesture from Bali ..." refers background in this paper
...To address weak classifier, the study by [16] has shown that a strong classifier can be built from relatively weak classifiers by building a cascaded classifier that forms a tree....
[...]
Book•
15 May 1999
TL;DR: In this article, the authors present a rigorous and complete textbook for a first course on information retrieval from the computer science (as opposed to a user-centred) perspective, which provides an up-to-date student oriented treatment of the subject.
Abstract: From the Publisher:
This is a rigorous and complete textbook for a first course on information retrieval from the computer science (as opposed to a user-centred) perspective. The advent of the Internet and the enormous increase in volume of electronically stored information generally has led to substantial work on IR from the computer science perspective - this book provides an up-to-date student oriented treatment of the subject.
9,923 citations
IBM1
TL;DR: An indexing method for time sequences for processing similarity queries using R * -trees to index the sequences and efficiently answer similarity queries and provides experimental results which show that the method is superior to search based on sequential scanning.
Abstract: We propose an indexing method for time sequences for processing similarity queries. We use the Discrete Fourier Transform (DFT) to map time sequences to the frequency domain, the crucial observation being that, for most sequences of practical interest, only the first few frequencies are strong. Another important observation is Parseval's theorem, which specifies that the Fourier transform preserves the Euclidean distance in the time or frequency domain. Having thus mapped sequences to a lower-dimensionality space by using only the first few Fourier coefficients, we use R * -trees to index the sequences and efficiently answer similarity queries. We provide experimental results which show that our method is superior to search based on sequential scanning. Our experiments show that a few coefficients (1–3) are adequate to provide good performance. The performance gain of our method increases with the number and length of sequences.
2,082 citations
"Grammar of Dance Gesture from Bali ..." refers methods in this paper
...Several algorithms have been proposed to simplify time series data such as Discrete Fourier Transformation [10], Discrete Wavelet Transformation [11], Piecewise Aggregate Approximation (PAA) [12], and Symbolic Aggregate Approximation (SAX) [13, 14]....
[...]
TL;DR: The cophenetic correlations which will be developed below provide an extremely simple and effective method for comparing dendrograms of various sorts.
Abstract: The purpose of this paper is to present a technique for comparing dendrograms resulting from numerical taxonomic research with one another and with dendrograms produced by conventional methods. One of the most frequent ways of depicting the results of studies in numerical taxonomy (Sokal, 1960; Sneath and Sokal, 1962) is by so-called dendrograms or diagrams of relationships. These are tree-like schemes which indicate the affinity of taxa to their nearest relatives (on the basis of similarity or phenetic resemblance alone, without any necessary phylogenetic implications). These diagrams resemble the customary phylogenetic trees, but are preferred for classificatory purposes; first, because phylogenetic inferences are speculative, while similarities are factual; secondly, because they are quantitative evaluations of these similarities; and thirdly, because they lack some of the other meanings often implied in phylogenetic trees (Sneath and Sokal, 1962). Such dendrograms have been published in bacteriological work (Sneath and Cowan, 1958), in studies of bees (Michener and Sokal 1957; Sokal and Michener 1958), butterflies (Ehrlich, 1961), rice (Morishima and Oka, 1960), members of the nightshade genus Solanum (Soria and Heiser, 1961) and others. With the increasing acceptance of the philosophy of numerical taxonomy an experimental phase in using various types of coefficients is beginning, which will involve the comparison of the results of numerical taxonomic research based on these different coefficients. So far we have lacked a procedure for such comparisons. The cophenetic correlations which will be developed below provide an extremely simple and effective method for comparing dendrograms of various sorts. Before proceeding to a detailed account of the technique, it will be useful to discuss briefly the four types of comparisons of dendrograms that we wish to make in numerical taxonomy and the reasons for them:
1,521 citations