Panacea+: Panoramic and Controllable Video Generation for Autonomous Driving
This work addresses the data scarcity problem for autonomous driving researchers and developers, though it is incremental as it builds upon previous work.
The paper tackles the need for high-quality annotated video training data in autonomous driving by proposing Panacea+, a framework for generating panoramic and controllable driving scene videos, which improves performance on tasks like 3D object tracking and detection across datasets such as nuScenes and Argoverse 2.
The field of autonomous driving increasingly demands high-quality annotated video training data. In this paper, we propose Panacea+, a powerful and universally applicable framework for generating video data in driving scenes. Built upon the foundation of our previous work, Panacea, Panacea+ adopts a multi-view appearance noise prior mechanism and a super-resolution module for enhanced consistency and increased resolution. Extensive experiments show that the generated video samples from Panacea+ greatly benefit a wide range of tasks on different datasets, including 3D object tracking, 3D object detection, and lane detection tasks on the nuScenes and Argoverse 2 dataset. These results strongly prove Panacea+ to be a valuable data generation framework for autonomous driving.