mlsanityData to 7 Oct 2026

FlowNet: Learning Optical Flow with Convolutional Networks

arXiv 1504.06852 ↗First posted 26 Apr 2015cs.CV cs.LGcited by 1,054

Not on the map. Why?

Cited by 41 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Convolutional neural networks (CNNs) have recently been very successful in a variety of computer vision tasks, especially on those linked to recognition. Optical flow estimation has not been among the tasks where CNNs were successful. In this paper we construct appropriate CNNs which are capable of solving the optical flow estimation problem as a supervised learning task. We propose and compare two architectures: a generic architecture and another one including a layer that correlates feature vectors at different image locations. Since existing ground truth data sets are not sufficiently large to train a CNN, we generate a synthetic Flying Chairs dataset. We show that networks trained on this unrealistic data still generalize very well to existing datasets such as Sintel and KITTI, achieving competitive accuracy at frame rates of 5 to 10 fps.

Key facts. Cited by 1,054 papers in the map (data through 7 Oct 2026). First posted on arXiv 26 Apr 2015.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 237; 2022: 205; 2023: 216; 2024: 170; 2025: 145; 2026: 81.

1,054 papers.

Sort:

By area:

  1. 2021Perceiver IO: A General Architecture for Structured Inputs & Outputscited by 616
  2. 2022BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detectioncited by 568
  3. 2023CoTracker: It is Better to Track Togethercited by 516
  4. 2022Kubric: A scalable dataset generatorcited by 373
  5. 2021RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matchingcited by 348
  6. 2022BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detectioncited by 340
  7. 2023Synthetic Data from Diffusion Models Improves ImageNet Classificationcited by 326
  8. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 324
  9. 2022Is synthetic data from generative models ready for image recognition?cited by 320
  10. 2023Unified-IO 2: Scaling Autoregressive Multimodal Models with Vision, Language, Audio, and Actioncited by 301
  11. 2023Human Preference Score: Better Aligning Text-to-Image Models with Human Preferencecited by 284
  12. 2023TAPIR: Tracking Any Point with per-frame Initialization and temporal Refinementcited by 281
38 more
  1. 2023PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Trackingcited by 274
  2. 2022FlowFormer: A Transformer Architecture for Optical Flowcited by 273
  3. 2022Unifying Flow, Stereo and Depth Estimationcited by 270
  4. 2022TAP-Vid: A Benchmark for Tracking Any Point in a Videocited by 258
  5. 2024Shape of Motion: 4D Reconstruction from a Single Videocited by 247
  6. 2025SAM 3D: 3Dfy Anything in Imagescited by 230
  7. 2021Learning to Estimate Hidden Motions with Global Motion Aggregationcited by 223
  8. 2022VRT: A Video Restoration Transformercited by 219
  9. 2023Drag Your GAN: Interactive Point-based Manipulation on the Generative Image Manifoldcited by 218
  10. 2022Particle Video Revisited: Tracking Through Occlusions Using Point Trajectoriescited by 215
  11. 2024Motion-I2V: Consistent and Controllable Image-to-Video Generation with Explicit Motion Modelingcited by 209
  12. 2021Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentationcited by 200
  13. 2022Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlationcited by 197
  14. 2021The Temporal Opportunist: Self-Supervised Multi-Frame Monocular Depthcited by 190
  15. 2023Tracking Everything Everywhere All at Oncecited by 190
  16. 2021COTR: Correspondence Transformer for Matching Across Imagescited by 182
  17. 2024SpatialTracker: Tracking Any 2D Pixels in 3D Spacecited by 178
  18. 2021Learning N:M Fine-grained Structured Sparse Neural Networks From Scratchcited by 177
  19. 2023StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learnerscited by 177
  20. 2024Motion Prompting: Controlling Video Generation with Motion Trajectoriescited by 176
  21. 2022Fake it till you make it: Learning transferable representations from synthetic ImageNet clonescited by 171
  22. 2022ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformercited by 170
  23. 2023Spring: A High-Resolution High-Detail Dataset and Benchmark for Scene Flow, Optical Flow and Stereocited by 170
  24. 2024SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flowcited by 169
  25. 2022Recurrent Video Restoration Transformer with Guided Deformable Attentioncited by 168
  26. 2023Think Twice before Driving: Towards Scalable Decoders for End-to-End Autonomous Drivingcited by 168
  27. 2022CroCo v2: Improved Cross-view Completion Pre-training for Stereo Matching and Optical Flowcited by 165
  28. 2022CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View Completioncited by 158
  29. 2021Occluded Video Instance Segmentation: A Benchmarkcited by 154
  30. 2021E-RAFT: Dense Optical Flow from Event Camerascited by 149
  31. 2023DDP: Diffusion Model for Dense Visual Predictioncited by 148
  32. 2023DynamicStereo: Consistent Dynamic Depth from Stereo Videoscited by 147
  33. 2022SHIFT: A Synthetic Driving Dataset for Continuous Multi-Task Domain Adaptationcited by 146
  34. 2021Parser-Free Virtual Try-on via Distilling Appearance Flowscited by 142
  35. 20214DComplete: Non-Rigid Motion Estimation Beyond the Observable Surfacecited by 123
  36. 2021Learning to Track with Object Permanencecited by 117
  37. 2023The Surprising Effectiveness of Diffusion Models for Optical Flow and Monocular Depth Estimationcited by 116
  38. 2023FlowFormer++: Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimationcited by 114

50 most cited of 1,054 All 1,054 (JSON)