JRDB-Reasoning: A Difficulty-Graded Benchmark for Visual Reasoning in Robotics
This provides a tailored benchmark for visual reasoning in human-crowded environments, addressing limitations in existing datasets for embodied AI agents, though it is incremental as it builds on the JRDB dataset.
The paper tackled the lack of clear complexity definitions and structured annotations in visual reasoning benchmarks for robotics by introducing JRDB-Reasoning, a difficulty-graded benchmark with adaptive query generation and step-by-step annotations, enabling fine-grained evaluation of models across reasoning levels.
Recent advances in Vision-Language Models (VLMs) and large language models (LLMs) have greatly enhanced visual reasoning, a key capability for embodied AI agents like robots. However, existing visual reasoning benchmarks often suffer from several limitations: they lack a clear definition of reasoning complexity, offer have no control to generate questions over varying difficulty and task customization, and fail to provide structured, step-by-step reasoning annotations (workflows). To bridge these gaps, we formalize reasoning complexity, introduce an adaptive query engine that generates customizable questions of varying complexity with detailed intermediate annotations, and extend the JRDB dataset with human-object interaction and geometric relationship annotations to create JRDB-Reasoning, a benchmark tailored for visual reasoning in human-crowded environments. Our engine and benchmark enable fine-grained evaluation of visual reasoning frameworks and dynamic assessment of visual-language models across reasoning levels.