PBench

'PBench' is a benchmark dataset that contains a diverse set of video-text pairs used to evaluate language-conditioned video generation models in the context of embodied intelligence tasks.

Papers using PBench (4)

Qwen-RobotWorld Technical Report: Unifying Embodied World Modeling through Language-Conditioned Video Generation2026EgoMind: Activating Spatial Cognition through Linguistic Reasoning in MLLMs2026SpaceMind: Camera-Guided Modality Fusion for Spatial Reasoning in Vision-Language Models2025Fpbench: A Comprehensive Benchmark Of Multimodal Large Language Models For Fingerprint Analysis2025