-
公开(公告)号：US12067661B2
公开(公告)日：2024-08-20
申请号：US17651330
申请日：2022-02-16
Applicant: Adobe Inc.
Inventor： Jun Saito , Nitin Saini , Ruben Villegas
CPC classification number: G06T13/40 , G06T9/001 , G06T13/205 , G06T17/00
Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing unsupervised learning of discrete human motions to generate digital human motion sequences. The disclosed system utilizes an encoder of a discretized motion model to extract a sequence of latent feature representations from a human motion sequence in an unlabeled digital scene. The disclosed system also determines sampling probabilities from the sequence of latent feature representations in connection with a codebook of discretized feature representations associated with human motions. The disclosed system converts the sequence of latent feature representations into a sequence of discretized feature representations by sampling from the codebook based on the sampling probabilities. Additionally, the disclosed system utilizes a decoder to reconstruct a human motion sequence from the sequence of discretized feature representations. The disclosed system also utilizes a reconstruction loss and a distribution loss to learn parameters of the discretized motion model.
-
2.发明公开公开(公告)号：US20240346737A1
公开(公告)日：2024-10-17
申请号：US18756135
申请日：2024-06-27
Applicant: Adobe Inc.
Inventor： Jun Saito , Nitin Saini , Ruben Villegas
CPC classification number: G06T13/40 , G06T9/001 , G06T13/205 , G06T17/00
Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing unsupervised learning of discrete human motions to generate digital human motion sequences. The disclosed system utilizes an encoder of a discretized motion model to extract a sequence of latent feature representations from a human motion sequence in an unlabeled digital scene. The disclosed system also determines sampling probabilities from the sequence of latent feature representations in connection with a codebook of discretized feature representations associated with human motions. The disclosed system converts the sequence of latent feature representations into a sequence of discretized feature representations by sampling from the codebook based on the sampling probabilities. Additionally, the disclosed system utilizes a decoder to reconstruct a human motion sequence from the sequence of discretized feature representations. The disclosed system also utilizes a reconstruction loss and a distribution loss to learn parameters of the discretized motion model.
-
3.发明公开公开(公告)号：US20230260182A1
公开(公告)日：2023-08-17
申请号：US17651330
申请日：2022-02-16
Applicant: Adobe Inc.
Inventor： Jun Saito , Nitin Saini , Ruben Villegas
CPC classification number: G06T13/40 , G06T13/205 , G06T9/001 , G06T17/00
Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for utilizing unsupervised learning of discrete human motions to generate digital human motion sequences. The disclosed system utilizes an encoder of a discretized motion model to extract a sequence of latent feature representations from a human motion sequence in an unlabeled digital scene. The disclosed system also determines sampling probabilities from the sequence of latent feature representations in connection with a codebook of discretized feature representations associated with human motions. The disclosed system converts the sequence of latent feature representations into a sequence of discretized feature representations by sampling from the codebook based on the sampling probabilities. Additionally, the disclosed system utilizes a decoder to reconstruct a human motion sequence from the sequence of discretized feature representations. The disclosed system also utilizes a reconstruction loss and a distribution loss to learn parameters of the discretized motion model.
-
-
Search Results
3
records
(took 0.01 seconds)
