File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1007/978-3-031-20074-8_19
- Scopus: eid_2-s2.0-85144492133
- Find via
Supplementary
-
Citations:
- Scopus: 2
- Appears in Collections:
Conference Paper: BRACE: The Breakdancing Competition Dataset for Dance Motion Synthesis
Title | BRACE: The Breakdancing Competition Dataset for Dance Motion Synthesis |
---|---|
Authors | |
Issue Date | 2022 |
Citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13668 LNCS, p. 329-344 How to Cite? |
Abstract | Generative models for audio-conditioned dance motion synthesis map music features to dance movements. Models are trained to associate motion patterns to audio patterns, usually without an explicit knowledge of the human body. This approach relies on a few assumptions: strong music-dance correlation, controlled motion data and relatively simple poses and movements. These characteristics are found in all existing datasets for dance motion synthesis, and indeed recent methods can achieve good results. We introduce a new dataset aiming to challenge these common assumptions, compiling a set of dynamic dance sequences displaying complex human poses. We focus on breakdancing which features acrobatic moves and tangled postures. We source our data from the Red Bull BC One competition videos. Estimating human keypoints from these videos is difficult due to the complexity of the dance, as well as the multiple moving cameras recording setup. We adopt a hybrid labelling pipeline leveraging deep estimation models as well as manual annotations to obtain good quality keypoint sequences at a reduced cost. Our efforts produced the BRACE dataset, which contains over 3 h and 30 min of densely annotated poses. We test state-of-the-art methods on BRACE, showing their limitations when evaluated on complex sequences. Our dataset can readily foster advance in dance motion synthesis. With intricate poses and swift movements, models are forced to go beyond learning a mapping between modalities and reason more effectively about body structure and movements. |
Persistent Identifier | http://hdl.handle.net/10722/352337 |
ISSN | 2023 SCImago Journal Rankings: 0.606 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Moltisanti, Davide | - |
dc.contributor.author | Wu, Jinyi | - |
dc.contributor.author | Dai, Bo | - |
dc.contributor.author | Loy, Chen Change | - |
dc.date.accessioned | 2024-12-16T03:58:20Z | - |
dc.date.available | 2024-12-16T03:58:20Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13668 LNCS, p. 329-344 | - |
dc.identifier.issn | 0302-9743 | - |
dc.identifier.uri | http://hdl.handle.net/10722/352337 | - |
dc.description.abstract | Generative models for audio-conditioned dance motion synthesis map music features to dance movements. Models are trained to associate motion patterns to audio patterns, usually without an explicit knowledge of the human body. This approach relies on a few assumptions: strong music-dance correlation, controlled motion data and relatively simple poses and movements. These characteristics are found in all existing datasets for dance motion synthesis, and indeed recent methods can achieve good results. We introduce a new dataset aiming to challenge these common assumptions, compiling a set of dynamic dance sequences displaying complex human poses. We focus on breakdancing which features acrobatic moves and tangled postures. We source our data from the Red Bull BC One competition videos. Estimating human keypoints from these videos is difficult due to the complexity of the dance, as well as the multiple moving cameras recording setup. We adopt a hybrid labelling pipeline leveraging deep estimation models as well as manual annotations to obtain good quality keypoint sequences at a reduced cost. Our efforts produced the BRACE dataset, which contains over 3 h and 30 min of densely annotated poses. We test state-of-the-art methods on BRACE, showing their limitations when evaluated on complex sequences. Our dataset can readily foster advance in dance motion synthesis. With intricate poses and swift movements, models are forced to go beyond learning a mapping between modalities and reason more effectively about body structure and movements. | - |
dc.language | eng | - |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | - |
dc.title | BRACE: The Breakdancing Competition Dataset for Dance Motion Synthesis | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1007/978-3-031-20074-8_19 | - |
dc.identifier.scopus | eid_2-s2.0-85144492133 | - |
dc.identifier.volume | 13668 LNCS | - |
dc.identifier.spage | 329 | - |
dc.identifier.epage | 344 | - |
dc.identifier.eissn | 1611-3349 | - |