Aerial–Marine Perception Dataset (AMP2026)
收藏资源简介:
AMP2026是由麦吉尔大学等机构联合构建的多平台海洋机器人数据集,旨在解决动态海洋环境下的跨视角追踪与地图构建难题。该数据集包含来自无人机、船载相机及水下机器人的同步传感数据,涵盖盐水和淡水两种环境,数据形式包括高分辨率影像、GNSS定位及惯性测量信息。通过标准化采集流程,数据集支持对目标追踪鲁棒性和多视角地图一致性的量化评估,主要服务于海洋自主系统、环境监测及跨模态感知算法开发。
AMP2026 is a multi-platform marine robotics dataset jointly constructed by McGill University and other institutions, aiming to address the challenges of cross-view tracking and map construction in dynamic marine environments. This dataset contains synchronized sensing data from unmanned aerial vehicles (UAVs), ship-borne cameras and underwater robots, covering both saltwater and freshwater environments. The data modalities include high-resolution imagery, GNSS positioning data and inertial measurement unit (IMU) information. Through a standardized data collection workflow, the dataset enables quantitative evaluation of target tracking robustness and multi-view map consistency, and primarily serves the development of marine autonomous systems, environmental monitoring and cross-modal perception algorithms.
数据集概述
基本信息
- 数据集名称: AMP2026
- 托管平台: Hugging Face
- 数据集地址: https://huggingface.co/datasets/edwinmeriaux/AMP2026
许可信息
- 许可证类型: Creative Commons Attribution 4.0 International (CC BY 4.0)
- 1AMP2026: A Multi-Platform Marine Robotics Dataset for Tracking and Mapping麦吉尔大学; MILA - 魁北克人工智能研究所; 巴黎萨克雷大学·中央高等电力学院; 明尼苏达大学; 德州农工大学; 伍兹霍尔海洋研究所; 宾汉姆顿大学 · 2026年



