DMD
收藏资源简介:
DMD数据集是由维科姆科技基金会创建的大型多模态驾驶监控数据集,专注于注意力与警觉性分析。该数据集包含41小时的RGB、深度和红外视频,涵盖37名司机在多种驾驶场景下的行为,如分心、视线分配、疲劳等。数据集设计旨在解决自动驾驶从SAE Level2向Level-3过渡中的关键问题,通过丰富的场景特征和多视角数据,推动驾驶监控系统的研究。此外,数据集还衍生出dBehaviourMD子集,用于深度学习训练,以实现实时、稳健的司机行为识别。
The DMD Dataset is a large-scale multimodal driving monitoring dataset created by the Vicome Technology Foundation, focusing on attention and alertness analysis. It contains 41 hours of RGB, depth and infrared videos, covering the driving behaviors of 37 drivers across various scenarios such as distraction, gaze allocation, fatigue and other related situations. The dataset is designed to address the core challenges during the transition of autonomous driving from SAE Level 2 to Level 3, and advance the research of driving monitoring systems through rich scene features and multi-view data. Moreover, the dataset also has a derivative subset named dBehaviourMD, which is utilized for deep learning training to achieve real-time and robust driver behavior recognition.

- 1DMD: A Large-Scale Multi-Modal Driver Monitoring Dataset for Attention and Alertness Analysis维科姆科技基金会 · 2020年



