Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385849319> ?p ?o ?g. }
Showing items 1 to 55 of
55
with 100 items per page.
- W4385849319 abstract "Recent advances in learning reusable motion priors have demonstrated their effectiveness in generating naturalistic behaviors. In this paper, we propose a new learning framework in this paradigm for controlling physics-based characters with significantly improved motion quality and diversity over existing state-of-the-art methods. The proposed method uses reinforcement learning (RL) to initially track and imitate life-like movements from unstructured motion clips using the discrete information bottleneck, as adopted in the Vector Quantized Variational AutoEncoder (VQ-VAE). This structure compresses the most relevant information from the motion clips into a compact yet informative latent space, i.e., a discrete space over vector quantized codes. By sampling codes in the space from a trained categorical prior distribution, high-quality life-like behaviors can be generated, similar to the usage of VQ-VAE in computer vision. Although this prior distribution can be trained with the supervision of the encoder's output, it follows the original motion clip distribution in the dataset and could lead to imbalanced behaviors in our setting. To address the issue, we further propose a technique named prior shifting to adjust the prior distribution using curiosity-driven RL. The outcome distribution is demonstrated to offer sufficient behavioral diversity and significantly facilitates upper-level policy learning for downstream tasks. We conduct comprehensive experiments using humanoid characters on two challenging downstream tasks, sword-shield striking and two-player boxing game. Our results demonstrate that the proposed framework is capable of controlling the character to perform considerably high-quality movements in terms of behavioral strategies, diversity, and realism. Videos, codes, and data are available at https://tencent-roboticsx.github.io/NCP/." @default.
- W4385849319 created "2023-08-16" @default.
- W4385849319 creator A5011329890 @default.
- W4385849319 creator A5019708503 @default.
- W4385849319 creator A5064069200 @default.
- W4385849319 creator A5071563180 @default.
- W4385849319 date "2023-08-14" @default.
- W4385849319 modified "2023-10-12" @default.
- W4385849319 title "Neural Categorical Priors for Physics-Based Character Control" @default.
- W4385849319 doi "https://doi.org/10.48550/arxiv.2308.07200" @default.
- W4385849319 hasPublicationYear "2023" @default.
- W4385849319 type Work @default.
- W4385849319 citedByCount "0" @default.
- W4385849319 crossrefType "posted-content" @default.
- W4385849319 hasAuthorship W4385849319A5011329890 @default.
- W4385849319 hasAuthorship W4385849319A5019708503 @default.
- W4385849319 hasAuthorship W4385849319A5064069200 @default.
- W4385849319 hasAuthorship W4385849319A5071563180 @default.
- W4385849319 hasBestOaLocation W43858493191 @default.
- W4385849319 hasConcept C101738243 @default.
- W4385849319 hasConcept C104114177 @default.
- W4385849319 hasConcept C107673813 @default.
- W4385849319 hasConcept C119857082 @default.
- W4385849319 hasConcept C154945302 @default.
- W4385849319 hasConcept C177769412 @default.
- W4385849319 hasConcept C41008148 @default.
- W4385849319 hasConcept C50644808 @default.
- W4385849319 hasConcept C5274069 @default.
- W4385849319 hasConcept C97541855 @default.
- W4385849319 hasConceptScore W4385849319C101738243 @default.
- W4385849319 hasConceptScore W4385849319C104114177 @default.
- W4385849319 hasConceptScore W4385849319C107673813 @default.
- W4385849319 hasConceptScore W4385849319C119857082 @default.
- W4385849319 hasConceptScore W4385849319C154945302 @default.
- W4385849319 hasConceptScore W4385849319C177769412 @default.
- W4385849319 hasConceptScore W4385849319C41008148 @default.
- W4385849319 hasConceptScore W4385849319C50644808 @default.
- W4385849319 hasConceptScore W4385849319C5274069 @default.
- W4385849319 hasConceptScore W4385849319C97541855 @default.
- W4385849319 hasLocation W43858493191 @default.
- W4385849319 hasOpenAccess W4385849319 @default.
- W4385849319 hasPrimaryLocation W43858493191 @default.
- W4385849319 hasRelatedWork W2159052453 @default.
- W4385849319 hasRelatedWork W2566616303 @default.
- W4385849319 hasRelatedWork W2734887215 @default.
- W4385849319 hasRelatedWork W2752972570 @default.
- W4385849319 hasRelatedWork W2803255133 @default.
- W4385849319 hasRelatedWork W2909431601 @default.
- W4385849319 hasRelatedWork W3013693939 @default.
- W4385849319 hasRelatedWork W3131327266 @default.
- W4385849319 hasRelatedWork W4294770367 @default.
- W4385849319 hasRelatedWork W4297051394 @default.
- W4385849319 isParatext "false" @default.
- W4385849319 isRetracted "false" @default.
- W4385849319 workType "article" @default.