Awesome-LiDAR-MOS

For anyone who wants to do research about 3D point cloud Moving Object Segmentation (MOS)

Continuously updating!

Benchmark

SemanticKITTI	MOE

Dataset

1. SemanticKITTI (From LMNet)
2. KITTI-Road (From MotionSeg3D)
3. Apollo (From MapMOS)
4. nuScenes (From MapMOS)
5. KITTI-Tracking-19 (From MapMOS)
6. SipailouCampus (From MotionBEV)
7. HeLiMOS (From HeLiMOS)
8. MOE-Dataset (From MOE)

Leader Board

This leaderboard features only open-source methods, allowing quick result reproduction.

• SemanticKITTI

Method	Val IoU[%]	Test IoU[%]
Auto-MOS	-	62.3
LMNet	67.1	62.5
4DMOS	71.9	65.2
MapMOS	86.1	66.0
MotionSeg3D	68.1	70.2
SFEMOS	-	71.2
AWV-MOS-LIO	70.0	-
MotionBEV	76.5	74.9
InsMOS	73.2	75.6
MF-MOS	76.1	76.7
CV-MOS	77.5	79.2
MambaMOS	82.3	80.1
4D-CS	80.9	83.5

• Apollo

All methods trained on the semanticKITTI and evaluated on the Apollo dataset

Method	Fine-tune	IoU[%]
MotionBEV	✓	61.4
LMNet	✓	65.9
MF-MOS	✓	70.7
MotionSeg3D	✓	72.3
4DMOS		73.1
CV-MOS	✓	75.8
InsMOS		78.0
MapMOS		81.2

Papers

-  Recent papers (from 2019)

Application: out.: outdoor   |   ind.: indoor  

Sensor: pc.: point cloud   |   img.: image   |   rad.: radar

Statistics: 🔥 code is available & stars >= 100  |  ⭐ citation >= 50

2019

• [CRV]Mapless Online Detection of Dynamic Objects in 3D Lidar.[out. pc.] ⭐

2020

• [IROS]Remove, then Revert: Static Point cloud Map Construction using Multiresolution Range Images.[code][out. pc.] 🔥 ⭐

2021

• [RAL]Moving Object Segmentation in 3D LiDAR Data: A Learning-based Approach Exploiting Sequential Data. [code][out. pc.] 🔥 ⭐
• [RAL]ERASOR: Egocentric Ratio of Pseudo Occupancy-Based Dynamic Object Removal for Static 3D Point Cloud Map Building. [code] [out. pc.] 🔥 ⭐
• [RAL]PointMoSeg: Sparse Tensor-Based End-to-End Moving-Obstacle Segmentation in 3-D Lidar Point Clouds for Autonomous Driving.[out. pc.]
• [ECMR]Mapping the Static Parts of Dynamic Scenes from 3D LiDAR Point Clouds Exploiting Ground Segmentation.[out. pc.]

2022

• [TPAMI]Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-Based Perception.[code] 🔥 ⭐
• [IROS]Efficient Spatial-Temporal Information Fusion for LiDAR-Based 3D Moving Object Segmentation.[code] [out. pc.] 🔥
• [IROS]PUA-MOS: End-to-End Point-wise Uncertainty Weighted Aggregation for Moving Object Segmentation.[code][out. pc.]
• [RAL]RVMOS: Range-View Moving Object Segmentation Leveraged by Semantic and Motion Features.[out. pc.]
• [RAL]Receding Moving Object Segmentation in 3D LiDAR Data Using Sparse 4D Convolutions.[code] [out. pc.] 🔥
• [WACV]Unsupervised 4D LiDAR Moving Object Segmentation in Stationary Settings with Multivariate Occupancy Time Series.[code] [out. pc.]
• [RAL]Automatic Labeling to Generate Training Data for Online LiDAR-Based Moving Object Segmentation.[code] [out. pc.] ⭐

2023

• [IROS]InsMOS: Instance-Aware Moving Object Segmentation in LiDAR Data.[code] [out. pc.] 🔥
• [IROS]LiDAR-SGMOS: Semantics-Guided Moving Object Segmentation with 3D LiDAR.[out. pc.]
• [RAL]Building Volumetric Beliefs for Dynamic Environments Exploiting Map-Based Moving Object Segmentation.[code] [out. pc.] 🔥
• [RAL]MotionBEV: Attention-Aware Online LiDAR Moving Object Segmentation With Bird’s Eye View Based Appearance and Motion Features.[code][out. pc.]
• [RAL]Dynablox: Real-Time Detection of Diverse Dynamic Objects in Complex Environments.[code][out. ind. pc.] 🔥
• [TRO]Radar Instance Transformer: Reliable Moving Instance Segmentation in Sparse Radar Point Clouds.[out. rad.]
• [ICCV]MarS3D: A Plug-and-Play Motion-Aware Model for Semantic Segmentation on Multi-Scan 3D Point Clouds.[code] [out. pc.]
• [Sensors]Real-Time LiDAR Point-Cloud Moving Object Segmentation for Autonomous Driving.[out. pc.]
• [IJAEOG]An efficient image-guided-based 3D point cloud moving object segmentation with transformer-attention in autonomous driving.[out. pc. img.]
• [TITS]3D-SeqMOS: A Novel Sequential 3D Moving Object Segmentation in Autonomous Driving.[out. pc.]
• [sensors]Real-Time LiDAR Point-Cloud Moving Object Segmentation for Autonomous Driving.[out. pc.]

2024

• [ICRA]MF-MOS: A Motion-Focused Model for Moving Object Segmentation.[code][out. pc.]
• [AAAI]RadarMOSEVE: A Spatial-Temporal Transformer Network for Radar-Only Moving Object Segmentation and Ego-Velocity Estimation.[code][out. rad.]
• [TASE]SegNet4D: Efficient Instance-Aware 4D Semantic Segmentation for LiDAR Point Cloud.[code][out. pc.]
• [TIM]CV-MOS: A Cross-View Model for Motion Segmentation.[code][out. pc.]
• [ACM MM]MambaMOS: LiDAR-based 3D Moving Object Segmentation with Motion-aware State Space Model.[code][out. pc.]
• [Nature Communications]Moving event detection from LiDAR point streams.[code][out. pc.]
• [TITS]Joint Scene Flow Estimation and Moving Object Segmentation on Rotational LiDAR Data.[code][out. pc.]
• [IROS]HeLiMOS: A Dataset for Moving Object Segmentation in 3D Point Clouds From Heterogeneous LiDAR Sensors.[code][out. pc.]
• [TIV]AWV-MOS-LIO: Adaptive Window Visibility based Moving Object Segmentation with LiDAR Inertial Odometry.[code][out. pc.]
• [IEEE Sens. J]Moving Object Segmentation Network for Multiview Fusion of Vehicle-Mounted LiDAR.[out. pc.]
• [IROS]MOE: A Dense LiDAR MOving Event Dataset, Detection Benchmark and LeaderBoard.[code][out. pc.]
• [TIM]SSF-MOS: Semantic Scene Flow Assisted Moving Object Segmentation for Autonomous Vehicles.[out. pc.]
• [RAL]4D-CS: Exploiting Cluster Prior for 4D Spatio-Temporal LiDAR Semantic Segmentation.[code][out. pc.]

2025

• [TIM]Real-Time Moving Object Detection for 3-D LiDAR Using Occlusion Accumulation in Range Image.[code][out. pc.]
• [IEEE IOT]FP-MOS: Frame-to-Frame Prediction for Dynamic Object Segmentation with LiDAR Data.[out. pc.]
• [Remote Sensing] Gradient Enhancement Techniques and Motion Consistency Constraints for Moving Object Segmentation in 3D LiDAR Point Clouds.[out. pc.]