Fast Neural Style Transfer for Motion Data

Automating motion style transfer can help save animators time by allowing them to produce a single set of motions, which can then be automatically adapted for use with different characters. The proposed fast, efficient technique for performing neural style transfer of human motion data uses a feed-f...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE computer graphics and applications Ročník 37; číslo 4; s. 42 - 49
Hlavní autoři: Holden, Daniel, Habibie, Ikhsanul, Kusajima, Ikuo, Komura, Taku
Médium: Magazine Article
Jazyk:angličtina
Vydáno: United States IEEE 2017
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Analytical models
Animation
computer graphics
Data models
Data transfer (computers)
deep learning
Human motion
machine learning
motion capture
Motion control
Neural networks
Optimization
style transfer
Training
Virtual reality
ISSN:0272-1716, 1558-1756, 1558-1756
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Automating motion style transfer can help save animators time by allowing them to produce a single set of motions, which can then be automatically adapted for use with different characters. The proposed fast, efficient technique for performing neural style transfer of human motion data uses a feed-forward neural network trained on a large motion database. The proposed framework can transform the style of motion thousands of times faster than previous approaches that use optimization.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0272-1716
1558-1756
1558-1756
DOI:10.1109/MCG.2017.3271464