dc.citation.conferencePlace |
CN |
- |
dc.citation.endPage |
5269 |
- |
dc.citation.startPage |
5261 |
- |
dc.citation.title |
AAAI Conference on Artificial Intelligence |
- |
dc.contributor.author |
Mascaró, Esteve Valls |
- |
dc.contributor.author |
Ahn, Hyemin |
- |
dc.contributor.author |
Lee, Dongheui |
- |
dc.date.accessioned |
2024-07-16T16:05:12Z |
- |
dc.date.available |
2024-07-16T16:05:12Z |
- |
dc.date.created |
2024-07-16 |
- |
dc.date.issued |
2024-02-25 |
- |
dc.description.abstract |
The synthesis of human motion has traditionally been addressed through task-dependent models that focus on specific challenges, such as predicting future motions or filling in intermediate poses conditioned on known key-poses. In this paper, we present a novel task-independent model called UNIMASK-M, which can effectively address these challenges using a unified architecture. Our model obtains comparable or better performance than the state-of-the-art in each field. Inspired by Vision Transformers (ViTs), our UNIMASK-M model decomposes a human pose into body parts to leverage the spatio-temporal relationships existing in human motion. Moreover, we reformulate various pose-conditioned motion synthesis tasks as a reconstruction problem with different masking patterns given as input. By explicitly informing our model about the masked joints, our UNIMASK-M becomes more robust to occlusions. Experimental results show that our model successfully forecasts human motion on the Human3.6M dataset while achieving state-of-the-art results in motion inbetweening on the LaFAN1 dataset for long transition periods. |
- |
dc.identifier.bibliographicCitation |
AAAI Conference on Artificial Intelligence, pp.5261 - 5269 |
- |
dc.identifier.doi |
10.1609/aaai.v38i6.28333 |
- |
dc.identifier.uri |
https://scholarworks.unist.ac.kr/handle/201301/83188 |
- |
dc.language |
영어 |
- |
dc.publisher |
Association for the Advancement of Artificial Intelligence |
- |
dc.title |
A Unified Masked Autoencoder with Patchified Skeletons for Motion Synthesis |
- |
dc.type |
Conference Paper |
- |
dc.date.conferenceDate |
2024-02-20 |
- |