Presentation
Stylized Text-to-Motion Generation via Hypernetwork-Driven Low-Rank Adaptation
SessionLearning in Motion
DescriptionWe propose a hypernetwork-driven LoRA framework for stylized text-to-motion generation that dynamically modulates a pretrained diffusion model using style embeddings extracted from reference motions. By structuring the style latent space with supervised contrastive learning and incorporating style-guided sampling, the method improves stylization quality, efficiency, and generalization to unseen motion styles.
Event Type
Technical Paper
TimeThursday, 23 July 20262:20pm - 2:30pm PDT
LocationRoom 403 A
Digital Library
PDF
Session Time & Location
Sunday, 19 July 20266:00pm - 8:45pm PDTHall K
Thursday, 23 July 20262:00pm - 3:30pm PDTRoom 403 A
Animation
Artificial Intelligence/Machine Learning
Simulation
Full Conference Supporter
Full Conference
