Semantic Map
Semantic maps represent environments by integrating spatial information with semantic labels, aiming to provide robots and autonomous systems with a richer understanding of their surroundings than traditional occupancy grids. Current research focuses on developing robust methods for creating and updating these maps using various sensor data (LiDAR, RGB-D cameras) and incorporating large language models (LLMs) for higher-level reasoning and instruction following. This work is significant because accurate and comprehensive semantic maps are crucial for enabling advanced capabilities in robotics, autonomous navigation, and remote sensing applications, such as improved object navigation, scene understanding, and instruction-following.
Papers
January 10, 2025
December 24, 2024
December 13, 2024
December 4, 2024
December 2, 2024
November 28, 2024
November 27, 2024
November 22, 2024
November 19, 2024
November 5, 2024
November 4, 2024
October 29, 2024
October 18, 2024
October 16, 2024
October 14, 2024
October 12, 2024
October 5, 2024
September 23, 2024