Max Robotics — Global Service Robot Directory + US Certification

About

Xihui Liu is a leading researcher at the forefront of embodied AI, 3D perception, and computer vision. Her work bridges the gap between visual understanding and language, enabling intelligent agents to perceive, reason about, and interact with complex 3D environments. She is best known for developing **EmbodiedScan** (2024, 54 citations), a holistic multi-modal 3D perception suite that equips embodied agents with the ability to fully understand scenes from first-person observations and contextualize them into language for interaction—a foundational contribution to embodied AI. Liu has also advanced open-vocabulary visual recognition with **OV-PARTS** (2023), enabling the segmentation of diverse object parts using arbitrary text descriptions, a critical capability for robotics and fine-grained scene understanding. Her work on **UniG3D** (2023) provides a unified dataset for 3D object generation, driving progress in generative AI for virtual reality and gaming. Additionally, **WorldSimBench** (2024) proposes a benchmark for evaluating video generation models as world simulators. With a rapidly growing citation impact and a focus on foundational, multi-modal perception, Xihui Liu is shaping the future of how machines see, understand, and simulate the world.