mlsanityData to 7 Oct 2026

Unsupervised Flow Refinement near Motion Boundaries

arXiv 2208.02305 ↗First posted 3 Aug 2022cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Unsupervised optical flow estimators based on deep learning have attracted increasing attention due to the cost and difficulty of annotating for ground truth. Although performance measured by average End-Point Error (EPE) has improved over the years, flow estimates are still poorer along motion boundaries (MBs), where the flow is not smooth, as is typically assumed, and where features computed by neural networks are contaminated by multiple motions. To improve flow in the unsupervised settings, we design a framework that detects MBs by analyzing visual changes along boundary candidates and replaces motions close to detections with motions farther away. Our proposed algorithm detects boundaries more accurately than a baseline method with the same inputs and can improve estimates from any flow predictor without additional training.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 30 of its 50 references that are in the map. First posted on arXiv 3 Aug 2022.

Builds on

30 of 50 references.

References · 30

  1. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  2. 2015FlowNet: Learning Optical Flow with Convolutional Networkscited by 1,054
  3. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 966
  4. 2016FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networkscited by 780
  5. 2016Optical Flow Estimation using a Spatial Pyramid Networkcited by 421
  6. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 324
  7. 2016One-Shot Video Object Segmentationcited by 233
  8. 20203D Photography using Context-aware Layered Depth Inpaintingcited by 206
  9. 2017UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional Census Losscited by 202
  10. 2019Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimationcited by 131
  11. 2019SelFlow: Self-Supervised Learning of Optical Flowcited by 128
  12. 2020Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimationcited by 120
18 more
  1. 2018MEMC-Net: Motion Estimation and Motion Compensation Driven Neural Network for Video Interpolation and Enhancementcited by 111
  2. 2019Dense Extreme Inception Network: Towards a Robust CNN Model for Edge Detectioncited by 104
  3. 2016Back to Basics: Unsupervised Learning of Optical Flow via Brightness Constancy and Motion Smoothnesscited by 101
  4. 2017Occlusion Aware Unsupervised Learning of Optical Flowcited by 101
  5. 2021Asymmetric Bilateral Motion Estimation for Video Frame Interpolationcited by 101
  6. 2017Video Object Segmentation Without Temporal Informationcited by 93
  7. 2022CRAFT: Cross-Attentional Flow Transformer for Robust Optical Flowcited by 82
  8. 2019A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularizationcited by 76
  9. 2021SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warpingcited by 74
  10. 2018Occlusions, Motion and Depth Boundaries with a Generic Network for Disparity, Optical Flow or Scene Flow Estimationcited by 69
  11. 2020UPFlow: Upsampling Pyramid for Unsupervised Optical Flow Learningcited by 49
  12. 2017On the Integration of Optical Flow and Action Recognitioncited by 42
  13. 2020LiteFlowNet3: Resolving Correspondence Ambiguity for More Accurate Optical Flow Estimationcited by 37
  14. 2018Deep Video Color Propagationcited by 21
  15. 2022Optical Flow Training under Limited Label Budget via Active Learningcited by 14
  16. 2015Coherent Motion Segmentation in Moving Camera Videos using Optical Flow Orientationscited by 8
  17. 2021Joint Detection of Motion Boundaries and Occlusionscited by 4
  18. 2017Boundary Flow: A Siamese Network that Predicts Boundary Motion without Training on Motioncited by 2

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 1.

3 papers.

  1. 2023SemARFlow: Injecting Semantics into Unsupervised Optical Flow Estimation for Autonomous Drivingcited by 9
  2. 2024UnSAMFlow: Unsupervised Optical Flow Guided by Segment Anything Modelcited by 8
  3. 2023UFD-PRiME: Unsupervised Joint Learning of Optical Flow and Stereo Depth through Pixel-Level Rigid Motion Estimationcited by 1