UnrealZoo Environments

该数据集包含了四个高保真的虚拟环境，旨在评估框架在具身视觉追踪方面的性能，尤其是在具有挑战性的条件下。这些环境包括：旧工厂、超市、地下停车场和化工厂，每个环境都设计来模拟不同的现实世界追踪难题。这一规模涵盖了四个各具特色的环境，任务是进行具身视觉追踪。

This dataset comprises four high-fidelity virtual environments intended to assess framework performance for embodied visual tracking, particularly under challenging real-world conditions. The included environments are old factories, supermarkets, underground parking lots, and chemical plants, each engineered to replicate distinct real-world tracking challenges. This dataset covers four unique environments, with the core task being embodied visual tracking.