WorldGuide is a framework that combines predictive pretraining on successful and failed trajectories with contrastive learning on matched success--failure pairs and provides a differentiable reward to guide joint optimization of the policy and visual encoder.
Lin Liu, Lu Zhang, Zi-Ying Song et al.· 0 citations
Long-horizon planning enables autonomous vehicles to anticipate scene evolution and potential risks, supporting safe and stable decisions in complex interactions. However, existing methods struggle to propagate motion trends from observed history into the future. Long rollouts based on a single latent state may further...
Zi-Ying Song, Sheng-Kai Zhang, Lei Yang et al.· 0 citations
VGGT provides a strong foundation for geometry-centric world models by recovering unified 3D scene geometry from visual observations. Although recent extensions enable temporal 3D prediction, their future evolution remains weakly conditioned on driving intentions and actions, limiting their ability to model alternative...
Zhao-Yang Liu, Kun Jiang, Zi-Ying Song et al.· 0 citations
Embodied world models predict the visual consequences of candidate actions before execution. However, existing action-conditioned world models often adopt uniformly weighted visual generation objectives that can be misaligned with embodied prediction needs. Even with explicit motion conditioning, these objectives can u...
Zi-Ying Song, Yu-Chen Liu, Zhuo-Ran Xu et al.· 0 citations
Joint-Embedding Predictive Architectures (JEPAs) provide a powerful framework for latent world modeling and planning in a reconstruction-free manner. Although numerous JEPA-based approaches have been proposed to mitigate representation collapse, our experiments on localized, out-of-distribution visual noise reveal that...
Reliable long-horizon planning remains a key challenge in end-to-end autonomous driving. By accounting for future motion evolution and potential consequences, it provides forward-looking guidance for safe and consistent driving in evolving traffic environments. Existing methods use historical planning states as tempora...
Yu-Chen Liu, Zi-Ying Song, Sheng-Kai Zhang et al.· 0 citations
PV-WM, a history-only world model over structured post-perception tracks, recurrently advances pedestrian root motion, 15-joint articulation, and learned vehicle states within a synchronized heterogeneous state while preserving type-specific pedestrian and vehicle dynamics.
Hao-Zhuang Chi, Jing-Song Liang, Zi-Ying Song et al.· 2 citations
MomADv2, a reliable state-space memory framework for long-horizon end-to-end autonomous driving, introduces a Selective State-Space Planning Memory Query Module, which filters historical planning queries based on temporal continuity and command consistency, and models the evolution of planning intentions through a sele...
Zi-Ying Song, Sheng-Kai Zhang, Lin Liu et al.· 0 citations
This work proposes Flow-JEPA (F-JEPA), a conditional flow matching dynamics model that jointly generates a sequence of future latent states conditioned on the current observation and actions, suggesting that conditional flow matching provides a promising alternative to deterministic autoregressive dynamics in JEPA worl...
PhyLatent, a dynamics-relevant training objective for JointEmbedding Predictive Architecture (JEPA) world models, shows that global non-collapse alone is insufficient for learning a reliable JEPA worldmodel state space.
Xi Zeng, Haojie Ren, Ziying Song· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
