📄 中文摘要
动态视觉传感器（DVS）的可用性限制对神经形态计算机视觉应用的研究者构成了根本性挑战。为此，研究社区创建了数据集，但通常样本或场景数量有限。为了解决神经形态视觉数据集缺乏全面仿真器的问题，提出了异常神经形态形状工具（ANTShapes），这是一种新颖的数据集仿真框架。ANTShapes基于Unity引擎构建，模拟了由展示随机生成行为的对象构成的抽象、可配置的三维场景，这些行为描述了运动和旋转等属性。对象行为的采样以及异常行为对象的标记为异常检测提供了新的可能性。
用于计算机视觉异常检测的神经形态数据集建模与仿真
📄 English Summary
Modelling and Simulation of Neuromorphic Datasets for Anomaly Detection in Computer Vision
The limitations on the availability of Dynamic Vision Sensors (DVS) pose a fundamental challenge for researchers in neuromorphic computer vision applications. In response, datasets have been created by the research community, but they often contain a limited number of samples or scenarios. To address the lack of a comprehensive simulator for neuromorphic vision datasets, the Anomalous Neuromorphic Tool for Shapes (ANTShapes) is introduced as a novel dataset simulation framework. Built on the Unity engine, ANTShapes simulates abstract, configurable 3D scenes populated by objects exhibiting randomly-generated behaviors that describe attributes such as motion and rotation. The sampling of object behaviors and the labeling of anomalously-acting objects offer new possibilities for anomaly detection.