# Consistency Guided Scene Flow Estimation

arXiv: https://arxiv.org/abs/2006.11242 · HTML: https://mlsanity.com/paper/2006.11242/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jun 2020.

## Built on by (most cited first, 5 of 5)

- [SCTN: Sparse Convolution-Transformer Network for Scene Flow Estimation](https://mlsanity.com/paper/2105.04447/) (2021)
- [PointFix: Learning to Fix Domain Bias for Robust Online Stereo Adaptation](https://mlsanity.com/paper/2207.13340/) (2022)
- [Learning Scene Flow in 3D Point Clouds with Noisy Pseudo Labels](https://mlsanity.com/paper/2203.12655/) (2022)
- [RMS-FlowNet++: Efficient and Robust Multi-Scale Scene Flow Estimation for Large-Scale Point Clouds](https://mlsanity.com/paper/2407.01129/) (2024)
- [EgoFlowNet: Non-Rigid Scene Flow from Point Clouds with Ego-Motion Support](https://mlsanity.com/paper/2407.02920/) (2024)

## Abstract (arXiv)

Consistency Guided Scene Flow Estimation (CGSF) is a self-supervised framework for the joint reconstruction of 3D scene structure and motion from stereo video. The model takes two temporal stereo pairs as input, and predicts disparity and scene flow. The model self-adapts at test time by iteratively refining its predictions. The refinement process is guided by a consistency loss, which combines stereo and temporal photo-consistency with a geometric term that couples disparity and 3D motion. To handle inherent modeling error in the consistency loss (e.g. Lambertian assumptions) and for better generalization, we further introduce a learned, output refinement network, which takes the initial predictions, the loss, and the gradient as input, and efficiently predicts a correlated output update. In multiple experiments, including ablation studies, we show that the proposed model can reliably predict disparity and scene flow in challenging imagery, achieves better generalization than the state-of-the-art, and adapts quickly and robustly to unseen domains.