边缘上的持续视觉异常检测：基准测试与高效解决方案
Continual Visual Anomaly Detection on the Edge: Benchmark and Efficient Solutions
AI总结：
本文提出边缘持续视觉异常检测的首个全面基准测试，评估了七种模型在三种轻量级架构上的表现，并提出Tiny-Dinomaly模型在内存和计算成本上显著降低的同时提升Pixel F1指标。
摘要：
视觉异常检测（VAD）是许多应用如工业检测和医疗领域的重要任务。尽管VAD已广泛研究，但两个关键挑战仍未得到充分解决：边缘部署，其中计算资源严重受限，以及持续学习，其中模型必须适应变化的数据分布而不遗忘先前知识。我们的基准测试为在联合效率和适应性约束下选择最优主干和VAD方法提供指导，表征内存足迹、推理成本和检测性能之间的权衡。单独研究这些挑战不足，因为为一种设置设计的方法在同时施加另一约束时会失效。本文提出首个边缘持续学习场景下的VAD基准测试，评估了七种VAD模型在三种轻量级主干架构上的表现。此外，我们提出Tiny-Dinomaly，一种基于DINO基础模型的轻量级Dinomaly变体，实现内存占用减少13倍、计算成本降低20倍，同时Pixel F1指标提升5个百分点。最后，我们对PatchCore和PaDiM进行针对性改进，以提高持续学习场景下的效率。
英文摘要：
Visual Anomaly Detection (VAD) is a critical task for many applications including industrial inspection and healthcare. While VAD has been extensively studied, two key challenges remain largely unaddressed in conjunction: edge deployment, where computational resources are severely constrained, and continual learning, where models must adapt to evolving data distributions without forgetting previously acquired knowledge. Our benchmark provides guidance for the selection of the optimal backbone and VAD method under joint efficiency and adaptability constraints, characterizing the trade-offs between memory footprint, inference cost, and detection performance. Studying these challenges in isolation is insufficient, as methods designed for one setting make assumptions that break down when the other constraint is simultaneously imposed. In this work, we propose the first comprehensive benchmark for VAD on the edge in the continual learning scenario, evaluating seven VAD models across three lightweight backbone architectures. Furthermore, we propose Tiny-Dinomaly, a lightweight adaptation of the Dinomaly model built on the DINO foundation model that achieves a 11$\times$ smaller memory footprint and 20$\times$ lower computational cost while improving localization (Pixel F1) by 5 percentage points. Finally, we introduce targeted modifications to PatchCore and PaDiM to improve their efficiency in the continual learning setting.
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。
