File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1145/2897824.2925975
- Scopus: eid_2-s2.0-84980028529
- WOS: WOS:000380112400108
- Find via
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: A deep learning framework for character motion synthesis and editing
Title | A deep learning framework for character motion synthesis and editing |
---|---|
Authors | |
Keywords | Human motion Convolutional neural networks Deep learning Manifold learning Autoencoder Character animation |
Issue Date | 2016 |
Citation | ACM Transactions on Graphics, 2016, v. 35, n. 4, article no. 138 How to Cite? |
Abstract | We present a framework to synthesize character movements based on high level parameters, such that the produced movements respect the manifold of human motion, trained on a large motion capture dataset. The learned motion manifold, which is represented by the hidden units of a convolutional autoencoder, represents motion data in sparse components which can be combined to produce a wide range of complex movements. To map from high level parametersto the motion manifold, we stack a deep feedforward neural network on top of the trained autoencoder. This network is trained to produce realistic motion sequences from parameters such as a curve over the terrain that the character should follow, or a target location for punching and kicking. The feedforward control network and the motion manifold are trained independently, allowing the user to easily switch between feedforward networks according to the desired interface, without re-training the motion manifold. Once motion is generated it can be edited by performing optimizationin the space of the motion manifold. This allows for imposing kinematic constraints, or transforming the style of the motion, while ensuring the edited motion remains natural. As a result, the system can produce smooth, high quality motion sequences without any manual pre-processing of the training data. |
Persistent Identifier | http://hdl.handle.net/10722/289058 |
ISSN | 2022 Impact Factor: 6.2 2020 SCImago Journal Rankings: 2.153 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Holden, Daniel | - |
dc.contributor.author | Saito, Jun | - |
dc.contributor.author | Komura, Taku | - |
dc.date.accessioned | 2020-10-12T08:06:34Z | - |
dc.date.available | 2020-10-12T08:06:34Z | - |
dc.date.issued | 2016 | - |
dc.identifier.citation | ACM Transactions on Graphics, 2016, v. 35, n. 4, article no. 138 | - |
dc.identifier.issn | 0730-0301 | - |
dc.identifier.uri | http://hdl.handle.net/10722/289058 | - |
dc.description.abstract | We present a framework to synthesize character movements based on high level parameters, such that the produced movements respect the manifold of human motion, trained on a large motion capture dataset. The learned motion manifold, which is represented by the hidden units of a convolutional autoencoder, represents motion data in sparse components which can be combined to produce a wide range of complex movements. To map from high level parametersto the motion manifold, we stack a deep feedforward neural network on top of the trained autoencoder. This network is trained to produce realistic motion sequences from parameters such as a curve over the terrain that the character should follow, or a target location for punching and kicking. The feedforward control network and the motion manifold are trained independently, allowing the user to easily switch between feedforward networks according to the desired interface, without re-training the motion manifold. Once motion is generated it can be edited by performing optimizationin the space of the motion manifold. This allows for imposing kinematic constraints, or transforming the style of the motion, while ensuring the edited motion remains natural. As a result, the system can produce smooth, high quality motion sequences without any manual pre-processing of the training data. | - |
dc.language | eng | - |
dc.relation.ispartof | ACM Transactions on Graphics | - |
dc.subject | Human motion | - |
dc.subject | Convolutional neural networks | - |
dc.subject | Deep learning | - |
dc.subject | Manifold learning | - |
dc.subject | Autoencoder | - |
dc.subject | Character animation | - |
dc.title | A deep learning framework for character motion synthesis and editing | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1145/2897824.2925975 | - |
dc.identifier.scopus | eid_2-s2.0-84980028529 | - |
dc.identifier.volume | 35 | - |
dc.identifier.issue | 4 | - |
dc.identifier.spage | article no. 138 | - |
dc.identifier.epage | article no. 138 | - |
dc.identifier.eissn | 1557-7368 | - |
dc.identifier.isi | WOS:000380112400108 | - |
dc.identifier.issnl | 0730-0301 | - |