全景深度预测

预测场景的语义与3D结构对于机器人安全导航与规划行动至关重要。近期方法已探索语义和全景场景预测，但未考虑场景的几何结构。本文提出全景深度预测任务，用于从单目相机图像预测未观测未来帧的全景分割与深度图。为支持本工作，我们扩展了流行的KITTI-360和Cityscapes基准数据集，通过从LiDAR点云计算深度图并利用顺序标注数据。我们还引入了一种合适的评估指标，协同量化预测的全景质量与深度估计精度。进一步，我们 presented two baselines and proposed the novel PDcast architecture that learns rich spatio-temporal representations by incorporating a transformer-based encoder, a forecasting module, and task-specific decoders to predict future panoptic-depth outputs. Extensive evaluations demonstrate the effectiveness of PDcast across two datasets and three forecasting tasks, consistently addressing the primary challenges. We make the code publicly available at https://pdcast.cs.uni-freiburg.de.