Targeted potential project (2023)

RC-DSAI, AI aided GNSS navigation, still working on the research problems

Abstract

Inspired

Dr. Wen, W. and Hu Runzhi.

Scholarship Plan

• (1st year) 09/2022-08/2023: H-ZGD2
• (2nd year) 09/2023-08/2024: H-ZGD2 - Reserved
• (3rd year) 09/2024-08/2025: H-ZGD2 - Project to be extended

Recent Research Plan: Hu Runzhi (Huawei ZGD2)

• Half year research plan (by Dec 2022)-Remote Sensing
  • Error quantification of the HD map generation (extrinsic and segmentation uncertainty)
  • System for HD generation based on multi-sensor integration
• One year research plan (by Jun 2023) IEEE ITSC 2023
  • Visual map update with crowdsourcing data (simulated dataset LVGI-input the HK scenario)
  • Error sources modeling of map update based on crowdsourcing data
• 1.5 years research plan (by Dec 2023)
  • Safety quantification of map update based on crowdsourcing data
  • Safety-constrained map update model
  • HD base map centered update monitoring and quantification

Student Supervision

• Supervise Edward? (undergraduate student)

• [12/01/2022]

  • Implement the code for the RTK positioning and image segmentation
  • Collect the data from the slope road to simulate the UAV scenarios dataset
  • Simulate the slop image to

News

• [12/01/2022]
  • Runzhi is working on the Meituan project where the vision is used to detect the GNSS NLOS. In particular, Runzhi is designing a UI to combine the AI for image segmentation and the RTKLIB for the GNSS-RTK positioning.
  • in the future, we can combine the GNSS-RTK FGO framework and replace the eixisting RTK posiitoning from the RTKLIB
• Discussion on 25th, Nov 2022
  • Dense matching with deep learning
  • AI for the GNSS-RTK positioning