SVIRO(Synthetic Vehicle Interior Rear Seat Occupancy Dataset)

SVIRO 是用于车辆内部后座占用检测和分类的综合数据集。该数据集由 10 辆不同车辆的 25.000 个风景组成，我们提供了几个模拟传感器输入和地面实况数据。在对有限数量的变体进行训练时，SVIRO 可用于评估机器学习模型的泛化能力和可靠性。我们的目标是在资源有限时提供一个通用基准来测试方法，这在工程应用中很常见。

SVIRO is a comprehensive dataset for in-vehicle rear seat occupancy detection and classification. This dataset comprises 25,000 scenes collected from 10 distinct vehicles, and we provide multiple simulated sensor inputs and ground truth data. When trained on a limited set of variations, SVIRO can be utilized to evaluate the generalization capability and reliability of machine learning models. Our goal is to offer a universal benchmark for testing methods under resource-constrained conditions, which is a common scenario in engineering applications.

搜集汇总

背景与挑战

背景概述

SVIRO是一个用于车辆内部后座占用检测和分类的综合数据集，包含10辆不同车辆的25,000个场景，提供模拟传感器输入和地面实况数据。该数据集旨在评估机器学习模型在有限训练变体下的泛化能力，可作为工程应用中的通用基准。