MambaFlow: A Novel and Flow-guided State Space Model for Scene Flow Estimation
Authors:
Preprint
Abstract
Scene flow estimation aims to predict 3D motion from consecutive point cloud frames, which is of great interest in autonomous driving field. Existing methods face challenges such as insufficient spatio-temporal modeling and inherent loss of fine-grained feature during voxelization. However, the success of Mamba, a representative state space model (SSM) that enables global modeling with linear complexity, provides a promising solution. In this paper, we propose MambaFlow, a novel scene flow estimation network with a mamba-based decoder. It enables deep interaction and coupling of spatio-temporal features using a well-designed backbone. Innovatively, we steer the global attention modeling of voxel-based features with point offset information using an efficient Mamba-based decoder, learning voxel-to-point patterns that are used to devoxelize shared voxel representations into point-wise features. To further enhance the model's generalization capabilities across diverse scenarios, we propose a novel scene-adaptive loss function that automatically adapts to different motion patterns.Extensive experiments on the Argoverse 2 benchmark demonstrate that MambaFlow achieves state-of-the-art performance with real-time inference speed among existing works, enabling accurate flow estimation in real-world urban scenarios. The code is available at https://github.com/SCNU-RISLAB/MambaFlow.Keywords
All Related Versions
This publication has 14 references indexed in Scilit:
- Flow4D: Leveraging 4D Voxel Network for LiDAR Scene Flow EstimationIEEE Robotics and Automation Letters, 2025
- SCIPaD: Incorporating Spatial Clues into Unsupervised Pose-Depth Joint LearningIEEE Transactions on Intelligent Vehicles, 2024
- AWV-MOS-LIO: Adaptive Window Visibility based Moving Object Segmentation with LiDAR Inertial OdometryIEEE Transactions on Intelligent Vehicles, 2024
- Beyond the Field-of-View: Enhancing Scene Visibility and Perception with Clip-Recurrent TransformerIEEE Transactions on Intelligent Vehicles, 2024
- RoadFormer: Duplex Transformer for RGB-Normal Semantic Road Scene ParsingIEEE Transactions on Intelligent Vehicles, 2024
- Augmenting Reinforcement Learning With Transformer-Based Scene Representation Learning for Decision-Making of Autonomous DrivingIEEE Transactions on Intelligent Vehicles, 2024
- Efficient 3D Object Detection Based on Pseudo-LiDAR RepresentationIEEE Transactions on Intelligent Vehicles, 2023
- FocusFlow: Boosting Key-Points Optical Flow Estimation for Autonomous DrivingIEEE Transactions on Intelligent Vehicles, 2023
- 3D Point-Voxel Correlation Fields for Scene Flow EstimationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
- Scalable Scene Flow From Point Clouds in the Real WorldIEEE Robotics and Automation Letters, 2021