mlsanityData to 14 Oct 2026

Towards Anytime Optical Flow Estimation with Event Cameras

arXiv 2307.05033 ↗First posted 11 Jul 2023cs.CV cs.RO eess.IVcited by 15

Not on the map. Why?

Abstract (arXiv)

Event cameras respond to changes in log-brightness at the millisecond level, making them ideal for optical flow estimation. However, existing datasets from event cameras provide only low frame rate ground truth for optical flow, limiting the research potential of event-driven optical flow. To address this challenge, we introduce a low-latency event representation, Unified Voxel Grid, and propose EVA-Flow, an EVent-based Anytime Flow estimation network to produce high-frame-rate event optical flow with only low-frame-rate optical flow ground truth for supervision. Furthermore, we propose the Rectified Flow Warp Loss (RFWL) for the unsupervised assessment of intermediate optical flow. A comprehensive variety of experiments on MVSEC, DESC, and our EVA-FlowSet demonstrates that EVA-Flow achieves competitive performance, super-low-latency (5ms), time-dense motion estimation (200Hz), and strong generalization. Our code will be available at https://github.com/Yaozhuwa/EVA-Flow.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 45 of its 70 references that are in the map. First posted on arXiv 11 Jul 2023.

Builds on

45 of 70 references.

References · 45

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,415
  3. 2015FlowNet: Learning Optical Flow with Convolutional Networkscited by 1,059
  4. 2019Event-based Vision: A Surveycited by 1,016
  5. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 968
  6. 2016FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networkscited by 784
  7. 2021DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Camerascited by 687
  8. 2016Optical Flow Estimation using a Spatial Pyramid Networkcited by 424
  9. 2017VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimatorcited by 423
  10. 2018Unsupervised Event-based Learning of Optical Flow, Depth, and Egomotioncited by 307
  11. 2021DSEC: A Stereo Event Camera Dataset for Driving Scenarioscited by 303
  12. 2022FlowFormer: A Transformer Architecture for Optical Flowcited by 275
33 more
  1. 2018The Multi Vehicle Stereo Event Camera Dataset: An Event Camera Dataset for 3D Perceptioncited by 254
  2. 2018EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based Camerascited by 224
  3. 2019End-to-End Learning of Representations for Asynchronous Event-Based Datacited by 217
  4. 2022Deep Patch Visual Odometrycited by 190
  5. 2016Deep Feature Flow for Video Recognitioncited by 186
  6. 2018A Unifying Contrast Maximization Framework for Event Cameras, with Applications to Motion, Depth, and Optical Flow Estimationcited by 178
  7. 2021E-RAFT: Dense Optical Flow from Event Camerascited by 152
  8. 2015Delving Deeper into Convolutional Networks for Learning Video Representationscited by 148
  9. 2021XVFI: eXtreme Video Frame Interpolationcited by 126
  10. 2021Self-Supervised Learning of Event-Based Optical Flow with Spiking Neural Networkscited by 99
  11. 2020Spike-FlowNet: Event-based Optical Flow Estimation with Energy-Efficient Hybrid Neural Networkscited by 98
  12. 2022Secrets of Event-Based Optical Flowcited by 92
  13. 2018DeepTAM: Deep Tracking and Mappingcited by 87
  14. 2019A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularizationcited by 76
  15. 2021Time-Ordered Recent Event (TORE) Volumes for Event Camerascited by 73
  16. 2017Semantic Video CNNs through Representation Warpingcited by 65
  17. 2022Learning Dense and Continuous Optical Flow from an Event Cameracited by 60
  18. 2021Spatio-Temporal Recurrent Networks for Event-Based Optical Flow Estimationcited by 50
  19. 2022Dense Continuous-Time Optical Flow from Events and Framescited by 46
  20. 2020ISSAFE: Improving Semantic Segmentation in Accidents by Fusing Event-based Datacited by 42
  21. 2023Taming Contrast Maximization for Learning Sequential, Low-latency, Event-based Optical Flowcited by 39
  22. 2023BlinkFlow: A Dataset to Push the Limits of Event-based Optical Flow Estimationcited by 33
  23. 2023TMA: Temporal Motion Aggregation for Event-based Optical Flowcited by 32
  24. 2022Lightweight Event-based Optical Flow Estimation via Iterative Deblurringcited by 31
  25. 2020Distance Surface for Event-Based Optical Flowcited by 29
  26. 2021Optical Flow Estimation for Spiking Cameracited by 29
  27. 2022PanoFlow: Learning 360deg Optical Flow for Surrounding Temporal Understandingcited by 26
  28. 2021Exploring Event-driven Dynamic Context for Accident Scene Segmentationcited by 23
  29. 2021Binary TTC: A Temporal Geofence for Autonomous Navigationcited by 20
  30. 2022CSFlow: Learning Optical Flow via Cross Strip Correlation for Autonomous Drivingcited by 16
  31. 2022Event-based Temporally Dense Optical Flow Estimation with Sequential Learningcited by 15
  32. 2016Can Ground Truth Label Propagation from Video help Semantic Segmentation?cited by 9
  33. 20223D-FlowNet: Event-based optical flow estimation with 3D representationcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 6; 2025: 4; 2026: 5.

15 papers.

Sort:

By area:

Similar, not cited

Close in wording, no citation either way

  1. 2026EventShiftFlow: Towards Hardware-efficient FPGA-based Flow Estimation
  2. 2025Spatio-Temporal State Space Model For Efficient Event-Based Optical Flow
  3. 2025Injecting Frame-Event Complementary Fusion into Diffusion for Optical Flow in Challenging Scenes
  4. 2025Perturbed State Space Feature Encoders for Optical Flow with Event Cameras
  5. 2026Bridging Event Streams and DiT: Event-Guided Video Frame Interpolation
  6. 2026TETO: Tracking Events with Teacher Observation for Motion Estimation and Frame Interpolation
  7. 2024An Event-Oriented Diffusion-Refinement Method for Sparse Events Completion
  8. 2025DESSERT: Diffusion-based Event-driven Single-frame Synthesis via Residual Training