Research
Research Directions研究方向
Our work explores intelligent IoT, low-power communication and sensing, and AI-assisted mobile systems across vehicles, drones, acoustic devices, and embodied robots. 我们的研究围绕智慧物联网、低功耗通信与感知，以及 AI 驱动的移动系统，覆盖车辆、无人机、声学设备和具身机器人等场景。
Intelligent Driving Sensing and Onboard mmWave Radar智能驾驶感知与车载毫米波雷达
We study how vehicles can sense occluded pedestrians and hidden hazards by combining onboard millimeter-wave radar, optical cues, and robust signal processing. The goal is to turn weak reflected paths into reliable perception for autonomous driving, ADAS, and cooperative road safety. 我们研究如何融合车载毫米波雷达、视觉线索与鲁棒信号处理，让车辆提前感知被遮挡的行人和潜在风险。目标是把微弱反射路径转化为可靠感知能力，服务于自动驾驶、辅助驾驶和车路协同安全。
UAV-Mounted LoRa Through-Wall Sensing and AI Life Localization无人机载 LoRa 穿墙感知与 AI 生命定位
We build lightweight UAV sensing platforms that use LoRa chirp signals, LiDAR, IMU, and edge AI to localize trapped people in emergency scenes. By fusing radio, geometry, and motion features, the system aims to output location heatmaps, life-state estimates, and confidence scores for rescue decisions. 我们构建轻量化无人机感知平台，利用 LoRa Chirp、LiDAR、IMU 与端侧 AI，在火灾等应急场景中定位室内被困人员。系统通过融合无线、几何和运动特征，输出位置热力图、生命状态和置信度，为救援决策提供依据。
Intelligent Acoustic Backscatter and AI Speech Recovery智能声学反向散射与 AI 语音恢复
We investigate acoustic backscatter systems that collect speech-related vibrations through low-power passive nodes. With physically guided neural recovery models, noisy backscatter signals can be enhanced and reconstructed for communication and sensing in challenging indoor environments. 我们研究低功耗被动节点采集语音相关振动的声学反向散射系统，并结合物理约束的神经网络恢复模型，对嘈杂反射信号进行增强与重建，用于复杂室内环境下的通信与感知。
Embodied Robot Mapping and Navigation具身机器人地图建模与导航
We explore bandwidth-aware mapping, multi-robot collaboration, and navigation for embodied intelligent systems. The research focuses on compact map representation, confidence-aware map fusion, and coordinated path planning for ground robots and aerial robots in complex indoor spaces. 我们探索面向具身智能系统的带宽感知地图建模、多机器人协同和导航技术，重点关注紧凑地图表示、置信度感知地图融合，以及地面机器人和空中机器人在复杂室内空间中的协同路径规划。