Active learning for on-road vehicle detection: a comparative study
Active learning for on-road vehicle detection: a comparative study
复制标题
DOI：
10.1007/s00138-011-0388-y
复制
发表时间：
2011-12
影响因子：
3.3
通讯作者：
Sayanan Sivaraman;Mohan M. Trivedi
中科院分区：
文献类型：
--
作者：
Sayanan Sivaraman;Mohan M. Trivedi
近年来，主动学习已成为利用计算机视觉构建稳健的目标检测系统的有力工具。事实上，主动学习方法在道路车辆检测中取得了令人印象深刻的成果。虽然在文献中已经探索和提出了用于目标检测的主动学习方法，但很少有研究来比较评估成本和优点。在这项研究中，我们对三种流行的用于道路车辆检测的主动学习方法进行了代价敏感分析。通过使用基于方向梯度特征直方图和支持向量机分类的检测器(HOG-SVM)和基于Haar-like特征和Adaboost分类的检测器(Haar-Adabost)进行的学习实验，证明了主动学习结果的普遍性。在静态图像和真实道路车辆数据集上进行了实验评估。学习方法是根据注释时间、所需数据、记忆力和精确度进行评估的。
In recent years, active learning has emerged as a powerful tool in building robust systems for object detection using computer vision. Indeed, active learning approaches to on-road vehicle detection have achieved impressive results. While active learning approaches for object detection have been explored and presented in the literature, few studies have been performed to comparatively assess costs and merits. In this study, we provide a cost-sensitive analysis of three popular active learning methods for on-road vehicle detection. The generality of active learning findings is demonstrated via learning experiments performed with detectors based on histogram of oriented gradient features and SVM classification (HOG–SVM), and Haar-like features and Adaboost classification (Haar–Adaboost). Experimental evaluation has been performed on static images and real-world on-road vehicle datasets. Learning approaches are assessed in terms of the time spent annotating, data required, recall, and precision.