mlsanityData to 7 Oct 2026

High-Resolution Optical Flow from 1D Attention and Correlation

arXiv 2104.13918 ↗First posted 28 Apr 2021cs.CVcited by 54

Not on the map. Why?

Abstract (arXiv)

Optical flow is inherently a 2D search problem, and thus the computational complexity grows quadratically with respect to the search window, making large displacements matching infeasible for high-resolution images. In this paper, we take inspiration from Transformers and propose a new method for high-resolution optical flow estimation with significantly less computation. Specifically, a 1D attention operation is first applied in the vertical direction of the target image, and then a simple 1D correlation in the horizontal direction of the attended image is able to achieve 2D correspondence modeling effect. The directions of attention and correlation can also be exchanged, resulting in two 3D cost volumes that are concatenated for optical flow estimation. The novel 1D formulation empowers our method to scale to very high-resolution input images while maintaining competitive performance. Extensive experiments on Sintel, KITTI and real-world 4K (2160 × 3840) resolution images demonstrated the effectiveness and superiority of our proposed method. Code and models are available at urlhttps://github.com/haofeixu/flow1d.

Key facts. Cited by 54 papers in the map (data through 7 Oct 2026). Builds on 28 of its 47 references that are in the map. First posted on arXiv 28 Apr 2021.

Builds on

28 of 47 references.

References · 28

  1. 2017Attention Is All You Needcited by 59,932
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2020End-to-End Object Detection with Transformerscited by 7,384
  4. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  5. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,830
  6. 2019Generating Long Sequences with Sparse Transformerscited by 1,523
  7. 2020Linformer: Self-Attention with Linear Complexitycited by 1,462
  8. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,303
  9. 2014Two-Stream Convolutional Networks for Action Recognition in Videoscited by 1,067
  10. 2015FlowNet: Learning Optical Flow with Convolutional Networkscited by 1,054
  11. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 966
  12. 2017The 2017 DAVIS Challenge on Video Object Segmentationcited by 942
16 more
  1. 2015A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimationcited by 878
  2. 2020Efficient Transformers: A Surveycited by 784
  3. 2016FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networkscited by 780
  4. 2018CCNet: Criss-Cross Attention for Semantic Segmentationcited by 717
  5. 2020Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentationcited by 326
  6. 2017Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolationcited by 321
  7. 2020AANet: Adaptive Aggregation Network for Efficient Stereo Matchingcited by 202
  8. 2019Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimationcited by 131
  9. 2019Learning the Depths of Moving People by Watching Frozen Peoplecited by 120
  10. 2020MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Maskcited by 117
  11. 2018Models Matter, So Does Training: An Empirical Study of CNNs for Optical Flow Estimationcited by 94
  12. 2018Hierarchical Discrete Distribution Decomposition for Match Density Estimationcited by 93
  13. 2015EpicFlow: Edge-Preserving Interpolation of Correspondences for Optical Flowcited by 80
  14. 2017Accurate Optical Flow via Direct Cost Volume Processingcited by 66
  15. 2016Full Flow: Optical Flow Estimation By Global Optimization over Regular Gridscited by 34
  16. 2018Devon: Deformable Volume Network for Learning Optical Flowcited by 16

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 14; 2023: 19; 2024: 13; 2025: 6; 2026: 1.

54 papers.

Sort:

By area:

  1. 2023CoTracker: It is Better to Track Togethercited by 516
  2. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 324
  3. 2022FlowFormer: A Transformer Architecture for Optical Flowcited by 273
  4. 2022Unifying Flow, Stereo and Depth Estimationcited by 270
  5. 2023AMT: All-Pairs Multi-Field Transforms for Efficient Frame Interpolationcited by 177
  6. 2023FlowFormer++: Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimationcited by 114
  7. 2023VideoFlow: Exploiting Temporal Cues for Multi-frame Optical Flow Estimationcited by 82
  8. 2023Motion-Guided Latent Diffusion for Temporally Consistent Real-world Video Super-resolutioncited by 59
  9. 2022SKFlow: Learning Optical Flow with Super Kernelscited by 56
  10. 2022Dense Continuous-Time Optical Flow from Events and Framescited by 45
  11. 2022Implicit Warping for Animation with Image Setscited by 45
  12. 2024TAPTR: Tracking Any Point with Transformers as Detectioncited by 43
38 more
  1. 2022DIP: Deep Inverse Patchmatch for High-Resolution Optical Flowcited by 40
  2. 2022Disentangling Architecture and Training for Optical Flowcited by 32
  3. 2023TransFlow: Transformer as Flow Learnercited by 32
  4. 2022Lightweight Event-based Optical Flow Estimation via Iterative Deblurringcited by 31
  5. 2023TMA: Temporal Motion Aggregation for Event-based Optical Flowcited by 31
  6. 2025DPFlow: Adaptive Optical Flow Estimation with a Dual-Pyramid Frameworkcited by 30
  7. 2022PanoFlow: Learning 360deg Optical Flow for Surrounding Temporal Understandingcited by 26
  8. 2023GAFlow: Incorporating Gaussian Attention into Optical Flowcited by 18
  9. 2024TAPTRv2: Attention-based Position Update Improves Tracking Any Pointcited by 18
  10. 2023AnyFlow: Arbitrary Scale Optical Flow with Implicit Neural Representationcited by 17
  11. 2022CSFlow: Learning Optical Flow via Cross Strip Correlation for Autonomous Drivingcited by 16
  12. 2023DistractFlow: Improving Optical Flow Estimation via Realistic Distractions and Pseudo-Labelingcited by 14
  13. 2024Track Everything Everywhere Fast and Robustlycited by 13
  14. 2022Self-supervised AutoFlowcited by 10
  15. 2022Multi-Scale RAFT: Combining Hierarchical Concepts for Learning-based Optical FLow Estimationcited by 9
  16. 2023SemARFlow: Injecting Semantics into Unsupervised Optical Flow Estimation for Autonomous Drivingcited by 9
  17. 2023Memory-Efficient Optical Flow via Radius-Distribution Orthogonal Cost Volumecited by 9
  18. 2024Rethinking RAFT for Efficient Optical Flowcited by 8
  19. 2024TAPTRv3: Spatial and Temporal Context Foster Robust Tracking of Any Point in Long Videocited by 8
  20. 2022GPS-GLASS: Learning Nighttime Semantic Segmentation Using Daytime Video and GPS datacited by 7
  21. 2023DIFT: Dynamic Iterative Field Transforms for Memory Efficient Optical Flowcited by 7
  22. 2023SplatFlow: Learning Multi-frame Optical Flow via Splattingcited by 7
  23. 2024NeuFlow: Real-time, High-accuracy Optical Flow Estimation on Robots Using Edge Devicescited by 7
  24. 2024Hybrid Cost Volume for Memory-Efficient Optical Flowcited by 6
  25. 2025MEMFOF: High-Resolution Training for Memory-Efficient Multi-Frame Optical Flow Estimationcited by 6
  26. 2023Unsupervised Optical Flow Estimation with Dynamic Timing Representation for Spike Cameracited by 5
  27. 2024Attention-aware non-rigid image registration for accelerated MR imagingcited by 4
  28. 2023FlowFormer: A Transformer Architecture and Its Masked Cost Volume Autoencoding for Optical Flowcited by 3
  29. 2025Removing Cost Volumes from Optical Flow Estimatorscited by 3
  30. 2023LLA-FLOW: A Lightweight Local Aggregation on Cost Volume for Optical Flow Estimationcited by 2
  31. 2024SciFlow: Empowering Lightweight Optical Flow Models with Self-Cleaning Iterationscited by 2
  32. 2024YOIO: You Only Iterate Once by mining and fusing multiple necessary global information in the optical flow estimationcited by 1
  33. 2024Deep-learning Optical Flow Outperforms PIV in Obtaining Velocity Fields from Active Nematicscited by 1
  34. 2025BAT: Learning Event-based Optical Flow with Bidirectional Adaptive Temporal Correlationcited by 1
  35. 2025Efficient All-Pairs Correlation Volume Sampling for Optical Flow Estimationcited by 1
  36. 2022CGCV:Context Guided Correlation Volume for Optical Flow Neural Networkscited by 0
  37. 2023Geometric Perception based Efficient Text Recognitioncited by 0
  38. 2023Context-Aware Iteration Policy Network for Efficient Optical Flow Estimationcited by 0

50 most cited of 54 All 54 (JSON)