arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.01328cs.CVcs.AIcs.LGcs.RO
收藏

视觉目标检测中开集误差的不确定度分析

Uncertainty for Identifying Open-Set Errors in Visual Object Detection

Queensland University of Technology (QUT)(昆士兰科技大学)

Dimity Miller, Niko Sünderhauf, Michael Milford, Feras Dayoub

摘要:

部署到开放世界中,对象检测器容易出现开集错误,即训练数据集中不存在的对象类的误报检测。我们提出了GMM-Det,一种实时的方法,用于从对象检测器中提取认知不确定性,以识别和拒绝开集错误。GMM-Det训练检测器产生一个结构化的logit空间,该空间使用特定于类的高斯混合模型进行建模。在测试时,开集错误通过它们在所有高斯混合模型下的低对数概率来识别。我们测试了两种常见的检测器架构,Faster R-CNN和RetinaNet,跨越机器人和计算机视觉的三个不同的数据集。我们的研究结果表明,GMM-Det始终优于现有的不确定性技术,用于识别和拒绝开集检测,特别是在安全关键应用所需的低错误率操作点。GMM-Det保持目标检测性能,并且仅引入最小的计算开销。我们还介绍了一种方法,用于将现有的目标检测数据集转换为特定的开集数据集,以评估开集性能的目标检测。

英文摘要:

Deployed into an open world, object detectors are prone to open-set errors, false positive detections of object classes not present in the training dataset. We propose GMM-Det, a real-time method for extracting epistemic uncertainty from object detectors to identify and reject open-set errors. GMM-Det trains the detector to produce a structured logit space that is modelled with class-specific Gaussian Mixture Models. At test time, open-set errors are identified by their low log-probability under all Gaussian Mixture Models. We test two common detector architectures, Faster R-CNN and RetinaNet, across three varied datasets spanning robotics and computer vision. Our results show that GMM-Det consistently outperforms existing uncertainty techniques for identifying and rejecting open-set detections, especially at the low-error-rate operating point required for safety-critical applications. GMM-Det maintains object detection performance, and introduces only minimal computational overhead. We also introduce a methodology for converting existing object detection datasets into specific open-set datasets to evaluate open-set performance in object detection.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑