CAP-DATA
收藏资源简介:
CAP-DATA数据集由长安大学交通学院的研究团队创建,包含11,727个交通事故视频,总计超过2.19百万帧。该数据集不仅标注了事故发生的时间窗口,还提供了详细的文本描述,包括事故前的实际情况、事故类别、事故原因和预防建议。数据集的创建旨在通过结合视觉和文本信息,提高交通事故预测的准确性和解释性,从而支持更安全的驾驶决策系统。
The CAP-DATA dataset was developed by a research team from the School of Transportation, Chang'an University. It contains 11,727 traffic accident videos, totaling over 2.19 million frames. Besides annotating the time windows of accident occurrences, it also provides detailed textual descriptions including the actual pre-accident scenario, accident category, accident cause and preventive suggestions. The dataset is created to improve the accuracy and interpretability of traffic accident prediction by combining visual and textual information, thus supporting safer driving decision-making systems.

- 1Cognitive Accident Prediction in Driving Scenes: A Multimodality Benchmark长安大学交通学院 · 2023年



