An audio-driven dancing avatar. Ofli, F., Demir, Y., Yemez, Y., Erzin, E., Tekalp, A. M., Balci, K., Kizoglu, I., Akarun, L., Canton-Ferrer, C., Tilmanne, J., Bozkurt, E., & Erdem, A. T. JOURNAL ON MULTIMODAL USER INTERFACES, 2(2, SI):93-103, SEP, 2008. doi abstract bibtex We present a framework for training and synthesis of an audio-driven dancing avatar. The avatar is trained for a given musical genre using the multicamera video recordings of a dance performance. The video is analyzed to capture the time-varying posture of the dancer's body whereas the musical audio signal is processed to extract the beat information. We consider two different marker-based schemes for the motion capture problem. The first scheme uses 3D joint positions to represent the body motion whereas the second uses joint angles. Body movements of the dancer are characterized by a set of recurring semantic motion patterns, i.e., dance figures. Each dance figure is modeled in a supervised manner with a set of HMM (Hidden Markov Model) structures and the associated beat frequency. In the synthesis phase, an audio signal of unknown musical type is first classified, within a time interval, into one of the genres that have been learnt in the analysis phase, based on mel frequency cepstral coefficients (MFCC). The motion parameters of the corresponding dance figures are then synthesized via the trained HMM structures in synchrony with the audio signal based on the estimated tempo information. Finally, the generated motion parameters, either the joint angles or the 3D joint positions of the body, are animated along with the musical audio using two different animation tools that we have developed. Experimental results demonstrate the effectiveness of the proposed framework.
@article{ ISI:000208536800003,
Author = {Ofli, Ferda and Demir, Yasemin and Yemez, Yucel and Erzin, Engin and
Tekalp, A. Murat and Balci, Koray and Kizoglu, Idil and Akarun, Lale and
Canton-Ferrer, Cristian and Tilmanne, Joelle and Bozkurt, Elif and
Erdem, A. Tanju},
Title = {{An audio-driven dancing avatar}},
Journal = {{JOURNAL ON MULTIMODAL USER INTERFACES}},
Year = {{2008}},
Volume = {{2}},
Number = {{2, SI}},
Pages = {{93-103}},
Month = {{SEP}},
Abstract = {{We present a framework for training and synthesis of an audio-driven
dancing avatar. The avatar is trained for a given musical genre using
the multicamera video recordings of a dance performance. The video is
analyzed to capture the time-varying posture of the dancer's body
whereas the musical audio signal is processed to extract the beat
information. We consider two different marker-based schemes for the
motion capture problem. The first scheme uses 3D joint positions to
represent the body motion whereas the second uses joint angles. Body
movements of the dancer are characterized by a set of recurring semantic
motion patterns, i.e., dance figures. Each dance figure is modeled in a
supervised manner with a set of HMM (Hidden Markov Model) structures and
the associated beat frequency. In the synthesis phase, an audio signal
of unknown musical type is first classified, within a time interval,
into one of the genres that have been learnt in the analysis phase,
based on mel frequency cepstral coefficients (MFCC). The motion
parameters of the corresponding dance figures are then synthesized via
the trained HMM structures in synchrony with the audio signal based on
the estimated tempo information. Finally, the generated motion
parameters, either the joint angles or the 3D joint positions of the
body, are animated along with the musical audio using two different
animation tools that we have developed. Experimental results demonstrate
the effectiveness of the proposed framework.}},
DOI = {{10.1007/s12193-008-0009-x}},
ISSN = {{1783-7677}},
ResearcherID-Numbers = {{Erzin, Engin/H-1716-2011}},
ORCID-Numbers = {{Erzin, Engin/0000-0002-2715-2368}},
Unique-ID = {{ISI:000208536800003}},
}
Downloads: 0
{"_id":"e3Kvmby2jPzf99Fw7","bibbaseid":"ofli-demir-yemez-erzin-tekalp-balci-kizoglu-akarun-etal-anaudiodrivendancingavatar-2008","downloads":0,"creationDate":"2016-12-09T13:41:59.858Z","title":"An audio-driven dancing avatar","author_short":["Ofli, F.","Demir, Y.","Yemez, Y.","Erzin, E.","Tekalp, A. M.","Balci, K.","Kizoglu, I.","Akarun, L.","Canton-Ferrer, C.","Tilmanne, J.","Bozkurt, E.","Erdem, A. T."],"year":2008,"bibtype":"article","biburl":"https://drive.google.com/uc?export=download&id=1d5Wvl98W_buJq6prXBz16a_6uF9Cdt-7","bibdata":{"bibtype":"article","type":"article","author":[{"propositions":[],"lastnames":["Ofli"],"firstnames":["Ferda"],"suffixes":[]},{"propositions":[],"lastnames":["Demir"],"firstnames":["Yasemin"],"suffixes":[]},{"propositions":[],"lastnames":["Yemez"],"firstnames":["Yucel"],"suffixes":[]},{"propositions":[],"lastnames":["Erzin"],"firstnames":["Engin"],"suffixes":[]},{"propositions":[],"lastnames":["Tekalp"],"firstnames":["A.","Murat"],"suffixes":[]},{"propositions":[],"lastnames":["Balci"],"firstnames":["Koray"],"suffixes":[]},{"propositions":[],"lastnames":["Kizoglu"],"firstnames":["Idil"],"suffixes":[]},{"propositions":[],"lastnames":["Akarun"],"firstnames":["Lale"],"suffixes":[]},{"propositions":[],"lastnames":["Canton-Ferrer"],"firstnames":["Cristian"],"suffixes":[]},{"propositions":[],"lastnames":["Tilmanne"],"firstnames":["Joelle"],"suffixes":[]},{"propositions":[],"lastnames":["Bozkurt"],"firstnames":["Elif"],"suffixes":[]},{"propositions":[],"lastnames":["Erdem"],"firstnames":["A.","Tanju"],"suffixes":[]}],"title":"An audio-driven dancing avatar","journal":"JOURNAL ON MULTIMODAL USER INTERFACES","year":"2008","volume":"2","number":"2, SI","pages":"93-103","month":"SEP","abstract":"We present a framework for training and synthesis of an audio-driven dancing avatar. The avatar is trained for a given musical genre using the multicamera video recordings of a dance performance. The video is analyzed to capture the time-varying posture of the dancer's body whereas the musical audio signal is processed to extract the beat information. We consider two different marker-based schemes for the motion capture problem. The first scheme uses 3D joint positions to represent the body motion whereas the second uses joint angles. Body movements of the dancer are characterized by a set of recurring semantic motion patterns, i.e., dance figures. Each dance figure is modeled in a supervised manner with a set of HMM (Hidden Markov Model) structures and the associated beat frequency. In the synthesis phase, an audio signal of unknown musical type is first classified, within a time interval, into one of the genres that have been learnt in the analysis phase, based on mel frequency cepstral coefficients (MFCC). The motion parameters of the corresponding dance figures are then synthesized via the trained HMM structures in synchrony with the audio signal based on the estimated tempo information. Finally, the generated motion parameters, either the joint angles or the 3D joint positions of the body, are animated along with the musical audio using two different animation tools that we have developed. Experimental results demonstrate the effectiveness of the proposed framework.","doi":"10.1007/s12193-008-0009-x","issn":"1783-7677","researcherid-numbers":"Erzin, Engin/H-1716-2011","orcid-numbers":"Erzin, Engin/0000-0002-2715-2368","unique-id":"ISI:000208536800003","bibtex":"@article{ ISI:000208536800003,\nAuthor = {Ofli, Ferda and Demir, Yasemin and Yemez, Yucel and Erzin, Engin and\n Tekalp, A. Murat and Balci, Koray and Kizoglu, Idil and Akarun, Lale and\n Canton-Ferrer, Cristian and Tilmanne, Joelle and Bozkurt, Elif and\n Erdem, A. Tanju},\nTitle = {{An audio-driven dancing avatar}},\nJournal = {{JOURNAL ON MULTIMODAL USER INTERFACES}},\nYear = {{2008}},\nVolume = {{2}},\nNumber = {{2, SI}},\nPages = {{93-103}},\nMonth = {{SEP}},\nAbstract = {{We present a framework for training and synthesis of an audio-driven\n dancing avatar. The avatar is trained for a given musical genre using\n the multicamera video recordings of a dance performance. The video is\n analyzed to capture the time-varying posture of the dancer's body\n whereas the musical audio signal is processed to extract the beat\n information. We consider two different marker-based schemes for the\n motion capture problem. The first scheme uses 3D joint positions to\n represent the body motion whereas the second uses joint angles. Body\n movements of the dancer are characterized by a set of recurring semantic\n motion patterns, i.e., dance figures. Each dance figure is modeled in a\n supervised manner with a set of HMM (Hidden Markov Model) structures and\n the associated beat frequency. In the synthesis phase, an audio signal\n of unknown musical type is first classified, within a time interval,\n into one of the genres that have been learnt in the analysis phase,\n based on mel frequency cepstral coefficients (MFCC). The motion\n parameters of the corresponding dance figures are then synthesized via\n the trained HMM structures in synchrony with the audio signal based on\n the estimated tempo information. Finally, the generated motion\n parameters, either the joint angles or the 3D joint positions of the\n body, are animated along with the musical audio using two different\n animation tools that we have developed. Experimental results demonstrate\n the effectiveness of the proposed framework.}},\nDOI = {{10.1007/s12193-008-0009-x}},\nISSN = {{1783-7677}},\nResearcherID-Numbers = {{Erzin, Engin/H-1716-2011}},\nORCID-Numbers = {{Erzin, Engin/0000-0002-2715-2368}},\nUnique-ID = {{ISI:000208536800003}},\n}\n\n","author_short":["Ofli, F.","Demir, Y.","Yemez, Y.","Erzin, E.","Tekalp, A. M.","Balci, K.","Kizoglu, I.","Akarun, L.","Canton-Ferrer, C.","Tilmanne, J.","Bozkurt, E.","Erdem, A. T."],"key":"ISI:000208536800003","id":"ISI:000208536800003","bibbaseid":"ofli-demir-yemez-erzin-tekalp-balci-kizoglu-akarun-etal-anaudiodrivendancingavatar-2008","role":"author","urls":{},"metadata":{"authorlinks":{"erzin, e":"http://home.ku.edu.tr/"}}},"search_terms":["audio","driven","dancing","avatar","ofli","demir","yemez","erzin","tekalp","balci","kizoglu","akarun","canton-ferrer","tilmanne","bozkurt","erdem"],"keywords":[],"authorIDs":["s4rze5RZET4EY5wXY"],"dataSources":["P7SB4qiBxZPhjXYRW","ziFHh7RJJaJNc9iie"]}