Papers
2
Total Citations
3
H-Index
1
About
Hongji Liu is a rising researcher in embodied intelligence and robotic perception, whose work bridges the critical gap between environmental mapping and multi-agent coordination. Liu’s primary research areas include dense panoptic mapping, hierarchical world representation, and multi-agent pathfinding (MAPF) for autonomous systems. In their landmark paper, “DHP-Mapping: A Dense Panoptic Mapping System with Hierarchical World Representation and Label Optimization Techniques” (2024, 2 citations), Liu introduced a novel framework that enables robots to access accurate, abstract-to-detailed geometric and semantic concepts from maps—a foundational contribution for informed decision-making in interactive tasks. This work addresses the longstanding challenge of efficiently modeling comprehensive environmental knowledge. Building on this, Liu’s 2025 paper, “GPU-accelerated Conflict-based Search for Multi-agent Embodied Intelligence” (1 citation), tackles the critical MAPF problem in dynamic environments, leveraging GPU acceleration to enhance coordination efficiency for applications ranging from autonomous robotics to smart transportation. Though early in their career, Liu’s innovative integration of mapping and multi-agent systems demonstrates significant potential to advance embodied intelligence, offering practical solutions for real-world autonomous navigation and coordination.
Research Focus
Key Achievements
Top Papers
- 1
- 2