Radar

• Here, the radar means Navtech radar data.
  • For details, see https://oxford-robotics-institute.github.io/radar-robotcar-dataset/
• Radar is robust to occlusions than LiDAR in urban sites.

Yeti Radar Odometry

• Paper
  • Do We Need to Compensate for Motion Distortion and Doppler Effects in Spinning Radar Navigation? (ICRA 2021)
• Code
  • https://github.com/keenan-burnett/yeti_radar_odometry
  • Features
    • implemented cen2018, cen2019 method with motion compensation
  • Requirements
    • need ray-wise timestamps

Tutorial: running Yeti on MulRan dataset

• see this video
  • link (TBA)

Results on MulRan dataset

• MulRan dataset
  • MulRan: Multimodal Range Dataset for Urban Place Recognition (ICRA 2019)
  • https://sites.google.com/view/mulran-pr/dataset
• Methods
  • Cen2018: Precise ego-motion estimation with millimeter-wave radar under diverse and challenging conditions (ICRA 2018)
  • Cen2019: Radar-only ego-motion estimation in difficult settings via graph matching (ICRA 2019)
• Result captures ^{1 2}
  • Left: Cen2018, Right: Cen2019
  • The trajectory color means start-blue and red-end.

Comments

Other posts

• [SLAM Study] » Gauss-Newton Opt == IEKF update ?
• [SLAM Study] » SLAM Back-end 공부자료 5개 추천
• [Radar SLAM] » ICRA 2021 Radar in Robotics Workshop 요약
• [LiDAR SLAM] » Scan Context-based LiDAR Pose-graph SLAM [구현]
• [Visual SLAM] » Filter-based VIO [1편] — MSCKF 계열 history 정리
• [Least Square Opt.] » Iterative Optimization [1편] — Nonlinear ICP 구현을 통해 알아보는 Gauss-newton Optimization
• [Filter-based SLAM] » Bayesian Filtering [2편] — Recursive estimation 의 시작 (칼만필터 유도)
• [Filter-based SLAM] » Bayesian Filtering [1편] — posterior 의 mean, covariance 구하기
• [Factor graph SLAM] » SLAM back-end [2++편] — Householder QR 분해 구현해보기
• [Factor graph SLAM] » SLAM back-end [2편] — Ax=b 대신 Ry=d 를 풀자