EnerVerse-AC: Envisioning Embodied Environments with Action Condition
This addresses the problem of expensive real-time interaction testing for robotic imitation learning researchers, though it appears incremental as it builds on prior architectures.
The paper tackles the high cost and challenge of testing robotic imitation learning in dynamic environments by proposing EnerVerse-AC, an action-conditional world model that generates future visual observations, which reduces costs while maintaining high fidelity in robotic manipulation evaluation.
Robotic imitation learning has advanced from solving static tasks to addressing dynamic interaction scenarios, but testing and evaluation remain costly and challenging due to the need for real-time interaction with dynamic environments. We propose EnerVerse-AC (EVAC), an action-conditional world model that generates future visual observations based on an agent's predicted actions, enabling realistic and controllable robotic inference. Building on prior architectures, EVAC introduces a multi-level action-conditioning mechanism and ray map encoding for dynamic multi-view image generation while expanding training data with diverse failure trajectories to improve generalization. As both a data engine and evaluator, EVAC augments human-collected trajectories into diverse datasets and generates realistic, action-conditioned video observations for policy testing, eliminating the need for physical robots or complex simulations. This approach significantly reduces costs while maintaining high fidelity in robotic manipulation evaluation. Extensive experiments validate the effectiveness of our method. Code, checkpoints, and datasets can be found at <https://annaj2178.github.io/EnerverseAC.github.io>.