用统计保证提升目标检测可靠性，让模型自知何时该信任自己。
Conformal Object Detection by Sequential Risk Control
- 通过序贯风险控制方法，为检测结果提供可验证的置信度
- 在多种场景下实现高精度与可靠性的平衡，满足安全关键需求
- 提供完整工具包，方便工业界复现和部署
近年来，目标检测器在工业应用中得到广泛采用，但在安全关键场景中的部署仍受限于神经网络固有的不可靠性以及目标检测模型的复杂结构。为此，本文引入共形预测（Conformal Prediction），一种具有统计保障的后处理不确定性量化方法，其有效性不依赖数据量大小或对模型/数据分布的先验知识。本文首次正式定义了共形目标检测（COD）问题，提出一种新方法——序贯共形风险控制（SeqCRC），将共形风险控制的统计保证扩展至两个序列任务，并适配两个参数，以满足COD需求。同时，我们设计了适用于不同场景与认证要求的新旧损失函数及预测集。最后，我们开发了一套共形工具包，支持方法复现与进一步探索。通过该工具包进行大量实验，验证了方法的有效性，并揭示了实际应用中的权衡与影响。
原文摘要 · Abstract (English)
Recent advances in object detectors have led to their adoption for industrial uses. However, their deployment in safety-critical applications is hindered by the inherent lack of reliability of neural networks and the complex structure of object detection models. To address these challenges, we turn to Conformal Prediction, a post-hoc predictive uncertainty quantification procedure with statistical guarantees that are valid for any dataset size, without requiring prior knowledge on the model or data distribution. Our contribution is manifold. First, we formally define the problem of Conformal Object Detection (COD). We introduce a novel method, Sequential Conformal Risk Control (SeqCRC), that extends the statistical guarantees of Conformal Risk Control to two sequential tasks with two parameters, as required in the COD setting. Then, we present old and new loss functions and prediction sets suited to applying SeqCRC to different cases and certification requirements. Finally, we present a conformal toolkit for replication and further exploration of our method. Using this toolkit, we perform extensive experiments that validate our approach and emphasize trade-offs and other practical consequences.
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。