Segmentation in the Wild

该数据集是一个由25个野外零样本分割数据集组成的基准，用于评估EfficientViT-SAM模型。此外，该基准还旨在评估模型在多样化、非结构化环境中的性能表现，其任务专注于真实世界场景下的零样本分割。

This dataset is a benchmark composed of 25 zero-shot segmentation datasets from outdoor real-world environments, which is intended for evaluating the EfficientViT-SAM model. Moreover, this benchmark aims to assess the model's performance in diverse and unstructured environments, with its task focusing on zero-shot segmentation in real-world scenarios.