# UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional Census Loss

arXiv: https://arxiv.org/abs/1711.07837 · HTML: https://mlsanity.com/paper/1711.07837/

## Key facts

- Cited by 202 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Nov 2017.

## Built on by (most cited first, 30 of 202)

- [GMFlow: Learning Optical Flow via Global Matching](https://mlsanity.com/paper/2111.13680/) (2021)
- [Unifying Flow, Stereo and Depth Estimation](https://mlsanity.com/paper/2211.05783/) (2022)
- [ProPainter: Improving Propagation and Transformer for Video Inpainting](https://mlsanity.com/paper/2309.03897/) (2023)
- [AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolation](https://mlsanity.com/paper/2304.09790/) (2023)
- [CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View Completion](https://mlsanity.com/paper/2210.10716/) (2022)
- [Upscale-A-Video: Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution](https://mlsanity.com/paper/2312.06640/) (2023)
- [CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos](https://mlsanity.com/paper/2410.11831/) (2024)
- [BootsTAP: Bootstrapped Training for Tracking-Any-Point](https://mlsanity.com/paper/2402.00847/) (2024)
- [IFRNet: Intermediate Feature Refine Network for Efficient Frame Interpolation](https://mlsanity.com/paper/2205.14620/) (2022)
- [Asymmetric Bilateral Motion Estimation for Video Frame Interpolation](https://mlsanity.com/paper/2108.06815/) (2021)
- [PDC-Net+: Enhanced Probabilistic Dense Correspondence Network](https://mlsanity.com/paper/2109.13912/) (2021)
- [FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis](https://mlsanity.com/paper/2312.17681/) (2023)
- [CityDreamer: Compositional Generative Model of Unbounded 3D Cities](https://mlsanity.com/paper/2309.00610/) (2023)
- [SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping](https://mlsanity.com/paper/2105.07014/) (2021)
- [Motion-Guided Latent Diffusion for Temporally Consistent Real-world Video Super-resolution](https://mlsanity.com/paper/2312.00853/) (2023)
- [Learning Monocular Depth in Dynamic Scenes via Instance-Aware Projection Consistency](https://mlsanity.com/paper/2102.02629/) (2021)
- [Representation Learning via Global Temporal Alignment and Cycle-Consistency](https://mlsanity.com/paper/2105.05217/) (2021)
- [Warp Consistency for Unsupervised Learning of Dense Correspondences](https://mlsanity.com/paper/2104.03308/) (2021)
- [Many-to-many Splatting for Efficient Video Frame Interpolation](https://mlsanity.com/paper/2204.03513/) (2022)
- [Drop-DTW: Aligning Common Signal Between Sequences While Dropping Outliers](https://mlsanity.com/paper/2108.11996/) (2021)
- [A Unified Pyramid Recurrent Network for Video Frame Interpolation](https://mlsanity.com/paper/2211.03456/) (2022)
- [Taming Contrast Maximization for Learning Sequential, Low-latency, Event-based Optical Flow](https://mlsanity.com/paper/2303.05214/) (2023)
- [RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust Correspondence Field Estimation and Pose Optimization](https://mlsanity.com/paper/2203.12870/) (2022)
- [Unifying Correspondence, Pose and NeRF for Pose-Free Novel View Synthesis from Stereo Pairs](https://mlsanity.com/paper/2312.07246/) (2023)
- [Learning Pixel Trajectories with Multiscale Contrastive Random Walks](https://mlsanity.com/paper/2201.08379/) (2022)
- [LRM-Zero: Training Large Reconstruction Models with Synthesized Data](https://mlsanity.com/paper/2406.09371/) (2024)
- [Disentangling Architecture and Training for Optical Flow](https://mlsanity.com/paper/2203.10712/) (2022)
- [Digging into Uncertainty in Self-supervised Multi-view Stereo](https://mlsanity.com/paper/2108.12966/) (2021)
- [Probabilistic Warp Consistency for Weakly-Supervised Semantic Correspondences](https://mlsanity.com/paper/2203.04279/) (2022)
- [Cross-View Completion Models are Zero-shot Correspondence Estimators](https://mlsanity.com/paper/2412.09072/) (2024)

## Abstract (arXiv)

In the era of end-to-end deep learning, many advances in computer vision are driven by large amounts of labeled data. In the optical flow setting, however, obtaining dense per-pixel ground truth for real scenes is difficult and thus such data is rare. Therefore, recent end-to-end convolutional networks for optical flow rely on synthetic datasets for supervision, but the domain mismatch between training and test scenarios continues to be a challenge. Inspired by classical energy-based optical flow methods, we design an unsupervised loss based on occlusion-aware bidirectional flow estimation and the robust census transform to circumvent the need for ground truth flow. On the KITTI benchmarks, our unsupervised approach outperforms previous unsupervised deep networks by a large margin, and is even more accurate than similar supervised methods trained on synthetic datasets alone. By optionally fine-tuning on the KITTI training data, our method achieves competitive optical flow accuracy on the KITTI 2012 and 2015 benchmarks, thus in addition enabling generic pre-training of supervised networks for datasets with limited amounts of ground truth.