无需标注数据，用图像增强一致性评估目标检测模型可靠性。
Cumulative Consensus Score: Label-Free and Model-Agnostic Evaluation of Object Detectors in Deployment
- 通过图像增强后框位置一致性计算共识得分。
- 在多个数据集上与真实指标相关性超90%。
- 适用于各类检测器，适合部署后持续监控。
在实际部署中评估目标检测模型面临标注缺失的挑战。本文提出累积共识得分（CCS），一种无需标注的监测信号，可用于实时环境中对检测器的持续评估与比较。方法对每张图像应用测试时数据增强，通过交并比衡量增强视图间预测框的空间一致性，所得共识得分可作为可靠性代理指标。在 Open Images 与 KITTI 的受控实验中，CCS 与 F1 分数、概率检测质量及最优修正成本的相关性超过 90%，在 COCO 与 BDD100K 上对多模型对的定性一致性也得到验证。该方法模型无关，适用于单阶段与两阶段检测器，且在案例级别运行，可识别性能下降场景。我们还提供了预期 CCS 与检测正确性的简化理论联系。总体而言，CCS 为检测器的 DevOps 式监控提供了可靠基础。
原文摘要 · Abstract (English)
Evaluating object detection models in deployment is challenging because ground-truth annotations are rarely available. We introduce the Cumulative Consensus Score (CCS), a label-free monitoring signal for continuous evaluation and comparison of detectors in real-world settings. CCS applies test-time data augmentation to each image and measures the spatial consistency of predicted bounding boxes across augmented views using Intersection over Union. The resulting consensus score serves as a proxy for reliability without requiring bounding box annotations. In controlled experiments on Open Images and KITTI, CCS achieved over 90% congruence with F1-score, Probabilistic Detection Quality, and Optimal Correction Cost, with qualitative consistency further confirmed on COCO and BDD100K across model pairs. The method is model-agnostic, working across single-stage and two-stage detectors, and operates at the case level to highlight under-performing scenarios. We also provide a simplified theoretical link between expected CCS and detection correctness. Altogether, CCS provides a robust foundation for DevOps-style monitoring of object detectors.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。