VLABench/vlm_evaluation_v1.0
收藏Hugging Face2025-03-28 更新2025-04-19 收录
下载链接：
https://hf-mirror.com/datasets/VLABench/vlm_evaluation_v1.0
下载链接
链接失效反馈官方服务：
资源简介：
VLABench评估数据集，用于评估视觉语言模型在具体化场景中的规划能力。包含CommonSense、Complex、M&T、PhysicsLaw、Semantic和Spatial等多个子任务，每个子任务包含100个数据集，提供视觉输入、任务描述和地面真实动作序列。
VLABench evaluation dataset designed to assess the planning capabilities of Vision-Language Models (VLMs) in embodied scenarios. It includes multiple subtasks such as CommonSense, Complex, M&T, PhysicsLaw, Semantic, and Spatial, each with 100 datasets providing visual input, task description, and ground-truth action sequence JSON files.
提供机构：
VLABench
