The paper introduces Structured Phase Space GAN (SPS‑GAN), a conditional generative model that learns structured dynamics from Cartesian trajectories and video. SPS‑GAN integrates a port‑Hamiltonian backbone with adversarial conditional generation, enabling representation of conservative, dissipative, and forced systems within a single architecture. Using a cyclic‑coordinate loss, the model recovers system degrees of freedom and canonical coordinates from raw observations, achieving better trajectory accuracy than supervised dynamics baselines and improved video consistency over generative baselines while consistently identifying correct latent mechanical coordinates.
arXiv:2607. 07763v1 Announce Type: new Abstract: World models are typically trained to predict discrete-time physical dynamics with a fixed step size baked into the model weights, preventing prediction at variable temporal resolutions.
arXiv:2605. 00412v3 Announce Type: replace Abstract: World models have recently re-emerged as a central paradigm for embodied intelligence, robotics, autonomous driving, and model-based reinforcement learning.
arXiv:2610.15267v1 Announce Type: cross Abstract: Joint-embedding predictive architectures (JEPAs) learn world models directly in latent space, but unconstrained autoregressive predictors may contrac...
arXiv:2505.15863v2 Announce Type: replace-cross Abstract: Generative AI (GenAI) is rapidly advancing the field of Autonomous Driving (AD), extending beyond traditional applications in text, image, an...
The paper introduces a layered evaluation protocol for generative scenario models used in autonomous driving, focusing on physical consistency and plausibility. It examines internal representations through kinematic alignment, statistical baseline comparison, latent controllability, and activation analysis, and then tests outputs against vehicle dynamics constraints such as lateral jerk thresholds. The protocol is applied to a VAE-based scenario generator and other generative models, revealing deeper insights than standard output-level metrics.
The paper proposes a continuous‑time generative framework that models time‑dependent data as evolution on a learned data manifold. By using pretrained score‑based models as geometric priors, it learns a vector field that drives data along score‑induced interpolation paths, enabling generation at arbitrary timestamps and temporal super‑resolution. The method includes a regression‑based training objective, a stability‑promoting term interpreted as denoising score matching, and a probabilistic extension for future trajectory distributions, demonstrated on natural video, PDE‑based fields, and molecular dynamics.
The paper introduces JoHan, a generative framework that directly recovers 2D and 3D hand motion from video sequences without intermediate per‑frame pose predictions. By jointly learning temporal dynamics and 2D‑3D correspondence, JoHan generates aligned pose sequences that improve temporal consistency and enable accurate estimation of the hand’s global position and orientation relative to the camera. Experiments on challenging benchmarks show that JoHan achieves higher accuracy and faster performance, producing smoother hand‑motion dynamics while maintaining high per‑frame pose accuracy.
arXiv:2602. 09303v2 Announce Type: replace Abstract: We propose a physics-informed consistency modeling framework for solving partial differential equations (PDEs) via fast, few-step generative inference.
arXiv:2512. 15313v2 Announce Type: replace-cross Abstract: Deep learning has become a standard approach for the modeling of audio effects, yet strictly black-box modeling remains problematic for time-varying systems.
arXiv:2607. 12143v1 Announce Type: cross Abstract: Accurately modeling the dynamics of planetary parachute and entry vehicle systems is critical for Entry, Descent, and Landing events such as vehicle separation and sensor activation.
arXiv:2606. 15015v1 Announce Type: cross Abstract: Physics-grounded video generation requires controllable 3D object dynamics that remain physically consistent under contact, deformation, and external forcing.
The paper introduces kernel-based methods for learning Hamiltonian systems directly from trajectory data, offering both a 2‑step approach (reconstruct trajectories first, then learn the Hamiltonian) and a 1‑step approach (joint inference). Experiments on mass‑spring dynamics, a nonlinear pendulum, and the Henon‑Heiles system show that the methods achieve accurate, data‑efficient predictions, outperforming 2‑step baselines especially when data are scarce, while preserving the Hamiltonian structure. The authors also provide a priori error estimates and a general numerical framework applicable to arbitrary dynamical systems.