Search papers, labs, and topics across Lattice.
This paper introduces a dual-stream autoregressive framework for modeling temporal cloth dynamics in animatable human avatars, addressing the limitations of existing methods that fail to capture realistic motion due to neglecting temporal causality. By integrating a geometric stream that propagates surface displacement with a state stream that incorporates historical states, the model achieves significant improvements in rendering quality and temporal coherence. Experiments show enhanced generalization to out-of-distribution poses, validating the effectiveness of explicitly modeling the temporal evolution of cloth dynamics.
Realistic cloth dynamics in animatable avatars can be achieved by explicitly modeling temporal causality, leading to significant improvements in rendering quality and generalization.
Creating photorealistic and temporally coherent animatable human avatars from RGB videos remains challenging. Current methods struggle to capture realistic cloth dynamics, producing over-smoothed appearance or severe artifacts on out-of-distribution poses. This limitation stems from a fundamental oversight: existing approaches neglect the temporal causality inherent in cloth physics, where current states emerge from previous states through temporal evolution rather than instantaneous skeletal configurations alone. Without explicit modeling of this causal structure, networks learn pose-appearance correlations instead of motion evolution, leading to poor generalization. We introduce a dual-stream autoregressive framework that explicitly models both observable geometric information and implicit internal state. The geometric stream propagates surface displacement from the previous frame, while the state stream fuses current features with historical states retrieved from a memory bank. Motion-adaptive aggregation handles spatially-varying dynamics, and adaptive regularization balances smoothness with flexibility. Experiments on challenging datasets demonstrate significant improvements in rendering quality, temporal consistency, and generalization to motion patterns beyond training distributions, validating that dual-stream temporal modeling enables realistic cloth dynamics.
