SLAM Phylogeny
Taxonomy Chapter map Handbook

SLAM Handbook decomposition

SLAM ?쇰Ц???꾪븳 MECE 怨꾪넻??/strong>

robotics-paper-phylogeny의 문법을 따르되, SLAM Handbook을 장별 목차가 아니라 시스템 기능 축으로 재배열했습니다. 논문 하나가 어디에 속하는지 판단하기 쉬운 4-depth tree입니다.

Current seed

8Phylum
31Class
64Order seeds
204Genus
1320References matched

Phylum relationship map

SLAM 시스템을 구성할 때 각 Phylum이 맡는 역할과 흐름

실선 화살표: SLAM main data flow 점선 화살표: 보조/피드백 관계 - 수학, 평가, learning, map cue는 여러 SLAM 모듈에 영향을 줌
SLAM Phylum relationship map Sensor and front-end produce measurements, back-end estimates state, map representations store the world, with modeling foundations below and spatial AI above. prediction cue: map-assisted tracking and relocalization Sensor & Odometry Visual/LiDAR/Radar/IMU Measurement Front-End Back-End Optimization Map Representations State, Geometry & Probabilistic Modeling Learning, Semantics & Spatial AI Problem & System Context Robustness, Evaluation & Operations

SLAM

왼쪽 트리에서 Phylum, Class, Order를 선택하면 해당 branch의 역할과 하위 항목이 표시됩니다.

Root 8 children

Children

Matched Handbook references How matched?

Primary label rules

  • sensor-specific full system은 Sensor & Odometry Modalities.
  • residual, registration, association, loop detection은 Measurement Front-End.
  • solver, inference, robust objective는 Back-End Optimization & Inference.
  • world representation은 Map Representations.
  • benchmark, metric, runtime, deployment는 Robustness, Evaluation & Operations.
  • learning 자체가 핵심 기여면 Learning, Semantics & Spatial AI.

Taxonomy mind map