Awesome Semantic-Object Mapping

This repository considers papers that optimize robot and object states based on Monocular, RGB-D, and LiDAR, as well as papers that perform loop closure detection through semantic scene matching.
🔥This repository is still under construction and will continue to be updated!🔥

Contents

• Papers
  • Object SLAM & Parameterization
  • Semantic Scene Matching for Loop Closing
• Resources
  • Datasets

News

24.08.01 First commit!

Papers

Object SLAM & Parameterization

Year	Venue	Paper Title	Repository
2024	arXiv	SlideSLAM: Sparse, Lightweight, Decentralized Metric-Semantic SLAM for Multi-Robot Navigation
2024	arXiv	Multi-Robot Object SLAM using Distributed Variational Inference
2024	ICRA	VOOM: Robust Visual Object Odometry and Mapping using Hierarchical Landmarks
2024	RA-L	ObVi-SLAM: Long-Term Object-Visual SLAM
2023	T-RO	An Object SLAM Framework for Association, Mapping, and High-Level Tasks	N/A
2023	T-ITS	Object SLAM With Robust Quadric Initialization and Mapping for Dynamic Outdoors	N/A
2023	RA-L	QISO-SLAM: Object-Oriented SLAM Using Dual Quadrics as Landmarks Based on Instance Segmentation	N/A
2023	JIRS [1]	SQ-SLAM: Monocular Semantic SLAM Based on Superquadric Object Representation
2023	ICRA	Object-based SLAM utilizing unambiguous pose parameters considering general symmetry types	N/A
2023	arXiv	UniQuadric: A SLAM Backend for Unknown Rigid Object 3D Tracking and Light-Weight Modeling	N/A
2022	ICRA	Unified Representation of Geometric Primitives for Graph-SLAM Optimization Using Decomposed Quadrics	N/A
2022	IROS	LayoutSLAM: Object Layout based Simultaneous Localization and Mapping for Reducing Object Map Distortion	N/A
2022	ISMAR	OA-SLAM: Leveraging Objects for Camera Relocalization in Visual SLAM
2022	RA-L	SO-SLAM: Semantic Object SLAM with Scale Proportional and Symmetrical Texture Constraints
2022	CVPR	Symmetry and Uncertainty-Aware Object SLAM for 6DoF Object Pose Estimation
2022	RA-L	Accurate and Robust Object SLAM With 3D Quadric Landmark Reconstruction in Outdoors	N/A
2021	IROS	LiDAR-Based Object-Level SLAM for Autonomous Vehicles	N/A
2021	ICCV	ODAM: Object Detection, Association, and Mapping using Posed RGB Video
2021	RA-L	DynaSLAM II: Tightly-Coupled Multi-Object Tracking and SLAM	N/A
2021	arXiv	Superquadric Object Representation for Optimization-based Semantic SLAM	N/A
2020	IROS	EAO-SLAM: Monocular Semi-Dense Object SLAM Based on Ensemble Data Association
2019	ICRA	Robust Object-based SLAM for High-speed Autonomous Navigation	N/A
2019	RA-L	Monocular Object and Plane SLAM in Structured Environments	N/A
2019	T-RO	CubeSLAM: Monocular 3D Object SLAM
2019	RA-L	QuadricSLAM: Dual Quadrics From Object Detections as Landmarks in Object-Oriented SLAM
2018	ACCV	Structure Aware SLAM using Quadrics and Planes	N/A

[1]: JIRS = Journal of Intelligent & Robotic Systems

[back to table]

Semantic Scene Matching

Year	Venue	Paper Title	Repository
2024	RA-L	SemanticTopoLoop: Semantic Loop Closure With 3D Topological Graph Based on Quadric-Level Object Map	N/A
2023	arXiv	Loop Closure Detection Based on Object-level Spatial Layout and Semantic Consistency	N/A
2022	RA-L	Towards Accurate Loop Closure Detection in Semantic SLAM With 3D Semantic Covisibility Graphs	N/A
2022	CASE	Object-based Loop Closure with Directional Histogram Descriptor
2021	JVCIR [2]	Semantic loop closure detection based on graph matching in multi-objects scenes	N/A
2021	RA-L	Topology Aware Object-Level Semantic Mapping Towards More Robust Loop Closure	N/A
2021	RA-L	Semantic histogram based graph matching for real-time multi-robot global localization in large scale environment
2019	ICRA	Global localization with objectlevel semantics and topology	N/A
2019	ICRA	Semantic Mapping for View-Invariant Relocalization	N/A
2017	RA-L	X-View: Graph-Based Semantic Multi-View Localization	N/A

[2]: JVCIR = Journal of Visual Communication and Image Representation

[back to table]

---

Resources

Datasets

Dataset Name	Description	Link
KITTI	A well-known dataset for autonomous driving research, offering stereo and LiDAR data for tasks like object detection, tracking, and SLAM.	KITTI
SemanticKITTI	A large-scale dataset providing semantic segmentation labels for LiDAR point clouds, based on the KITTI Odometry Benchmark, suitable for autonomous driving applications.	SemanticKITTI
Replica	A highly detailed and realistic dataset of 18 indoor environments, providing photorealistic textures and semantic annotations for 3D research.	Replica
TUM	A dataset for evaluating RGB-D SLAM systems, containing benchmark sequences recorded with a Microsoft Kinect camera.	TUM
Matterport3D	A large dataset of indoor RGB-D images with 3D meshes and semantic annotations, used for developing scene understanding algorithms.	Matterport3D
SUN RGB-D	A dataset containing RGB-D images from various sensors, annotated with 3D bounding boxes for object detection and scene understanding research.	SUN RGB-D
Microsoft-RGBD	A dataset featuring RGB-D sequences captured in various indoor environments, used for scene understanding and 3D reconstruction tasks.	Microsoft-RGBD
NYU Depth v2	An RGB-D dataset collected with a Kinect sensor, providing labeled depth maps and semantic annotations for indoor scenes.	NYU Depth v2
Redwood-OS	A collection of RGB-D sequences designed for evaluating 3D reconstruction algorithms, with a focus on object and scene scanning.	Redwood-OS
RGBD Scenes v1	A dataset with RGB-D video sequences for studying 3D object recognition and scene understanding, featuring cluttered indoor environments.	RGBD Scenes v1
RGBD Scenes v2	An extended version of RGBD Scenes v1, providing additional objects and sequences for more complex scene analysis.	RGBD Scenes v2
ICL-NUIM	A synthetic RGB-D dataset for benchmarking SLAM and visual odometry algorithms, featuring photo-realistic scenes.	ICL-NUIM
ScanNet	A richly annotated dataset with RGB-D video sequences of indoor environments, aimed at advancing 3D scene understanding and reconstruction.	ScanNet
YCB-Video	A dataset for benchmarking 6-DoF object pose estimation in video, including real and synthetic sequences of household objects.	YCB-Video
ApolloScape	A comprehensive dataset for autonomous driving, providing annotated 2D/3D images and point clouds from real-world traffic scenarios.	ApolloScape
Cityscapes	A large-scale dataset for semantic urban scene understanding, featuring pixel-level annotations for various urban street scenes.	Cityscapes

[back to table]