[ECCV 2024] ViewFormer: Exploring Spatiotemporal Modeling for Multi-View 3D Occupancy Perception via View-Guided Transformers
-
Updated
Nov 1, 2024 - Python
[ECCV 2024] ViewFormer: Exploring Spatiotemporal Modeling for Multi-View 3D Occupancy Perception via View-Guided Transformers
Physics-informed, RL-aligned evaluation engine for autonomous driving: JAX/Flax NNX trajectory-diffusion world models on the Waymo Open Dataset, DPO and reward-guided adversarial steering toward the scenarios that break AV stacks, physics feasibility, occupancy flow, WOSAC-style realism metrics and sensor simulation.
Analyze intersection motion from overhead imagery with trajectories, velocity, TTC, uncertainty, and collision-risk counterfactuals.
To associate your repository with the occupancy-flow topic, visit your repo's landing page and select "manage topics."