Home Research Papers Datasets State of the Art Pricing

Discover, visualize, and connect AI research papers. Explore the latest trends and insights in artificial intelligence research.

Product

Home
Research Papers
About

Support

Contact
Terms of Service
Privacy Policy

© 2026 Papersgraph. All rights reserved.

computer-vision-2

Self-supervised Scene Flow Estimation

3260 papers • 126 benchmarks • 313 datasets

Self-supervised method only with lidar point clouds as input to predict flows of each point. Self-supervised way for scene flow estimation Public leaderboard: Argoverse 2.0 Scene Flow Argoverse 2 2024 Scene Flow Challenge

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in self-supervised-scene-flow-estimation-2

Trend

Dataset

Best Model

Actions

Argoverse 2

Argoverse 2

Libraries

i

Use these libraries to find self-supervised-scene-flow-estimation-2 models and implementations

Datasets

Argoverse 2

Subtasks

No subtasks available.

Most implemented papers

Self-Supervised Scene Flow Estimation With 4-D Automotive Radar

Chris Xiaoxuan Lu, Fangqiang Ding, Zhijun Pan, Yimin Deng, Jianning Deng•Tue Mar 01 2022

This work aims to address the above challenges and estimate scene flow from 4-D radar point clouds by leveraging self-supervised learning and a robust scene flow estimation architecture and three novel losses are bespoken designed to cope with intractable radar data.

51

Content

Introduction Benchmarks Datasets Subtasks Libraries Papers

0

PointPWC-Net: A Coarse-to-Fine Network for Supervised and Self-Supervised Scene Flow Estimation on 3D Point Clouds

Wenxuan Wu, Fuxin Li, Zhuwen Li, Wei Liu, Zhiyuan Wang•Tue Nov 26 2019

This work proposes a novel end-to-end deep scene flow model, called PointPWC-Net, on 3D point clouds in a coarse- to-fine fashion, which shows great generalization ability on KITTI Scene Flow 2015 dataset, outperforming all previous methods.

87 0

Just Go With the Flow: Self-Supervised Scene Flow Estimation

David Held, Himangi Mittal, Brian Okorn•Sat Nov 30 2019

This work presents a method of training scene flow that uses two self-supervised losses, based on nearest neighbors and cycle consistency, which matches current state-of-the-art supervised performance using no real world annotations and exceeds state- of- the-art performance when combining the self- supervised approach with supervised learning on a smaller labeled dataset.

162 0

Adversarial Self-Supervised Scene Flow Estimation

Victor Zuanazzi, Joris van Vugt, O. Booij, Pascal Mettes•Sat Oct 31 2020

This work proposes a metric learning approach for self-supervised scene flow estimation, where a network learns a latent metric to distinguish between points translated by flow estimations and the target point cloud.

10 0

FlowStep3D: Model Unrolling for Self-Supervised Scene Flow Estimation

Yonina C. Eldar, D. Raviv, Yair Kittenplon•Wed Nov 18 2020

This work presents a recurrent architecture that learns a single step of an unrolled iterative alignment procedure for refining scene flow predictions, and demonstrates iterative convergence toward the solution using strong regularization.

125 0

Fast Neural Scene Flow

S. Lucey, J. K. Pontes, Xueqian Li, Jianqiao Zheng, Francesco Ferroni•Mon Apr 17 2023

The fast neural scene flow (FNSF) approach reports for the first time real-time performance comparable to learning methods, without any training or OOD bias on two of the largest open autonomous driving (AV) lidar datasets Waymo Open and Argoverse.

40 0

Occlusion Guided Self-supervised Scene Flow Estimation on 3D Point Clouds

D. Raviv, Bojun Ouyang•Fri Apr 09 2021

This work presents a new self-supervised training method and an architecture for the 3D scene flow estimation under occlusions that outperforms traditional architectures by a large margin for occluded and non-occluded scenarios.

19 0

3D Object Detection with a Self-supervised Lidar Scene Flow Backbone

Alois Knoll, Ming-Yu Liu, Ekim Yurtsever, Zhijie Yang, Emecc Erccelik, Hanzhen Zhang, Pinar Topccam, Maximilian Listl, Yilmaz Kaan cCayli•Sun May 01 2022

This work combines a self-supervised backbone with a supervised 3D detection head model that learns to utilize motion representations to distinguish dynamic objects exhibiting different movement patterns and shows the relationship between self- supervised multi-frame flow representations and single-frame3D detection hypotheses.

34 0

ZeroFlow: Scalable Scene Flow via Distillation

James Hays, Deva Ramanan, Dinesh Jayaraman, Kyle Vedder, Neehar Peri, Nathaniel Chodosh, Ishan Khatri, Eric Eaton, Yang Liu, Georgia Tech•Tue May 16 2023

This work proposes Scene Flow via Distillation, a simple, scalable distillation framework that uses a label-free optimization method to produce pseudo-labels to supervise a feedforward model, and achieves state-of-the-art performance on the Argoverse 2 Self-Supervised Scene Flow Challenge while using zero human labels.

12 0

Neural Scene Flow Prior

S. Lucey, J. K. Pontes, Xueqian Li•Sun Oct 31 2021

This paper revisits the scene flow problem that relies predominantly on runtime optimization and strong regularization and includes the inclusion of a neural scene flow prior, which uses the architecture of neural networks as a new type of implicit regularizer.

121 0

Adding a benchmark result helps the community track progress.

Self-supervised Scene Flow Estimation | State-of-the-Art