智能网联汽车数据集

通过网约车运营，基于车载智能终端设备、智驾量产车采集、轻量化加装等方式，实现规模化、低成本的智驾数据采集和测试，结合多年积累，形成极具价值的数据集。涵盖：舱内舱外各类场景库、感知数据集、规控数据集、事故数据集、VLM语义场景集、地图库等多模态数据集，涉及视频、图片、语音、点云等，数据维度丰富，以供学术科研、智能驾驶、智慧交通等场景数据服务。This highly valuable dataset is created via large-scale and low-cost intelligent driving data collection and testing, which is carried out through ride-hailing operations using in-vehicle smart terminal devices, mass-produced intelligent driving vehicles and lightweight retrofitting approaches, combined with years of accumulated industry data and experience. It encompasses multimodal datasets including various in-cabin and out-of-cabin scenario libraries, perception datasets, planning and control datasets, accident datasets, VLM semantic scenario sets and map databases. These datasets cover multiple data modalities such as video, image, audio, point cloud and others, feature rich data dimensions, and can offer data services for academic research, intelligent driving, intelligent transportation and other application scenarios.

搜集汇总

背景与挑战

背景概述

该数据集通过网约车运营、车载智能终端设备等多种方式实现规模化、低成本的智能驾驶数据采集，涵盖舱内舱外场景库、感知数据集、规控数据集等多模态内容，包括视频、图片、语音和点云等丰富数据维度。它旨在为学术科研、智能驾驶和智慧交通等领域提供全面的数据服务支持，具有广泛的应用价值。