提升驾驶行为分析的可靠性，让模型判断更准且可信。
Calibrated and Resource-Aware Super-Resolution for Reliable Driver Behavior Analysis
- 动态调整超分辨率策略，兼顾计算资源与模型校准。
- 关键事件检测准确率超基线，瞌睡识别AUPR达0.78。
- 适合对安全性要求极高的自动驾驶场景使用。
驾驶监控系统不仅需要高精度，还必须具备可靠的置信度评分以保障安全。直接在低分辨率数据上训练虽能获得高整体准确率，但预测结果校准性差，在安全关键场景中可能带来风险。本文提出一种资源感知的自适应超分辨率框架，优化模型校准性及关键事件的精确率-召回率表现。该方法在安全相关指标上达到当前最佳：校准误差（ECE）为5.8%（低于基线6.2%），瞌睡检测的AUPR达0.78（基线0.74），手机使用检测的精确率-召回率组合达0.74（基线0.71）。引入轻量级伪影检测器（0.3M参数，5.2ms开销），有效过滤超分辨率带来的幻觉。尽管低分辨率训练模型是通用强基线，但本框架是安全关键应用中可靠性的最优解。
原文摘要 · Abstract (English)
Driver monitoring systems require not just high accuracy but reliable, well-calibrated confidence scores for safety-critical deployment. While direct low-resolution training yields high overall accuracy, it produces poorly calibrated predictions that can be dangerous in safety-critical scenarios. We propose a resource-aware adaptive super-resolution framework that optimizes for model calibration and high precision-recall on critical events. Our approach achieves state-of-the-art performance on safety-centric metrics: best calibration (ECE of 5.8\% vs 6.2\% for LR-trained baselines), highest AUPR for drowsiness detection (0.78 vs 0.74), and superior precision-recall for phone use detection (0.74 vs 0.71). A lightweight artifact detector (0.3M parameters, 5.2ms overhead) provides additional safety by filtering SR-induced hallucinations. While LR-trained video models serve as strong general-purpose baselines, our adaptive framework represents the state-of-the-art solution for safety-critical applications where reliability is paramount.
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。