Mixed Signals
收藏资源简介:
Mixed Signals数据集是由康奈尔大学等机构创建的高质量、大规模的V2X协同感知数据集,包含45.1k个点云和240.6k个边界框标注。数据集由三个装有不同类型LiDAR传感器的自动驾驶车辆以及一个装有双LiDAR传感器的路边单元共同采集,覆盖了10个不同类别,包括4类脆弱道路使用者。该数据集是首个包含异构自动驾驶车辆LiDAR配置和左侧驾驶国家(澳大利亚)场景的V2X数据集,为自动驾驶车辆的协同感知研究提供了精确对齐和一致标注的高质量数据。
The Mixed Signals Dataset is a high-quality, large-scale V2X cooperative perception dataset developed by Cornell University and other institutions. It contains 45.1k point clouds and 240.6k annotated bounding boxes. The dataset was collected via three autonomous vehicles equipped with different types of LiDAR sensors and a roadside unit fitted with dual LiDAR sensors. It covers 10 distinct categories, including four types of vulnerable road users. Notably, this is the first V2X dataset that incorporates heterogeneous autonomous vehicle LiDAR configurations and scenarios from left-hand drive countries (Australia), providing high-quality, precisely aligned and consistently annotated data for autonomous vehicle cooperative perception research.

- 1Mixed Signals: A Diverse Point Cloud Dataset for Heterogeneous LiDAR V2X Collaboration康奈尔大学 · 2025年



