mlsanityData to 14 Oct 2026

BAT: Learning Event-based Optical Flow with Bidirectional Adaptive Temporal Correlation

arXiv 2503.03256 ↗First posted 5 Mar 2025cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

Event cameras deliver visual information characterized by a high dynamic range and high temporal resolution, offering significant advantages in estimating optical flow for complex lighting conditions and fast-moving objects. Current advanced optical flow methods for event cameras largely adopt established image-based frameworks. However, the spatial sparsity of event data limits their performance. In this paper, we present BAT, an innovative framework that estimates event-based optical flow using bidirectional adaptive temporal correlation. BAT includes three novel designs: 1) a bidirectional temporal correlation that transforms bidirectional temporally dense motion cues into spatially dense ones, enabling accurate and spatially dense optical flow estimation; 2) an adaptive temporal sampling strategy for maintaining temporal consistency in correlation; 3) spatially adaptive temporal motion aggregation to efficiently and adaptively aggregate consistent target motion features into adjacent motion features while suppressing inconsistent ones. Our results rank 1st on the DSEC-Flow benchmark, outperforming existing state-of-the-art methods by a large margin while also exhibiting sharp edges and high-quality details. Notably, our BAT can accurately predict future optical flow using only past events, significantly outperforming E-RAFT's warm-start approach. Code: textcolormagentahttps://github.com/gangweiX/BAT.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 44 of its 52 references that are in the map. First posted on arXiv 5 Mar 2025.

Builds on

44 of 52 references.

References · 44

  1. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,415
  2. 2015FlowNet: Learning Optical Flow with Convolutional Networkscited by 1,059
  3. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 968
  4. 2016FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networkscited by 784
  5. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 327
  6. 2018Unsupervised Event-based Learning of Optical Flow, Depth, and Egomotioncited by 307
  7. 2021DSEC: A Stereo Event Camera Dataset for Driving Scenarioscited by 303
  8. 2022Vision Transformer with Deformable Attentioncited by 295
  9. 2022FlowFormer: A Transformer Architecture for Optical Flowcited by 275
  10. 2018The Multi Vehicle Stereo Event Camera Dataset: An Event Camera Dataset for 3D Perceptioncited by 254
  11. 2018EV-FlowNet: Self-Supervised Optical Flow Estimation for Event-based Camerascited by 224
  12. 2021Learning to Estimate Hidden Motions with Global Motion Aggregationcited by 224
32 more
  1. 2017UnFlow: Unsupervised Learning of Optical Flow with a Bidirectional Census Losscited by 202
  2. 2024SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flowcited by 174
  3. 2023Spring: A High-Resolution High-Detail Dataset and Benchmark for Scene Flow, Optical Flow and Stereocited by 172
  4. 2022CroCo v2: Improved Cross-view Completion Pre-training for Stereo Matching and Optical Flowcited by 166
  5. 2021E-RAFT: Dense Optical Flow from Event Camerascited by 152
  6. 2021TimeLens: Event-based Video Frame Interpolationcited by 151
  7. 2020Back to Event Basics: Self-Supervised Learning of Image Reconstruction for Event Cameras via Photometric Constancycited by 125
  8. 2023The Surprising Effectiveness of Diffusion Models for Optical Flow and Monocular Depth Estimationcited by 116
  9. 2022Time Lens++: Event-based Frame Interpolation with Parametric Non-linear Flow and Multi-scale Fusioncited by 111
  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. 2018Models Matter, So Does Training: An Empirical Study of CNNs for Optical Flow Estimationcited by 94
  13. 2022Secrets of Event-Based Optical Flowcited by 92
  14. 2022CRAFT: Cross-Attentional Flow Transformer for Robust Optical Flowcited by 82
  15. 2023VideoFlow: Exploiting Temporal Cues for Multi-frame Optical Flow Estimationcited by 82
  16. 2019A Lightweight Optical Flow CNN - Revisiting Data Fidelity and Regularizationcited by 76
  17. 2022Learning Dense and Continuous Optical Flow from an Event Cameracited by 60
  18. 2022SKFlow: Learning Optical Flow with Super Kernelscited by 56
  19. 2024MemFlow: Optical Flow Estimation and Prediction with Memorycited by 55
  20. 2021High-Resolution Optical Flow from 1D Attention and Correlationcited by 54
  21. 2021Learning Optical Flow from a Few Matchescited by 53
  22. 2021Spatio-Temporal Recurrent Networks for Event-Based Optical Flow Estimationcited by 50
  23. 2023Taming Contrast Maximization for Learning Sequential, Low-latency, Event-based Optical Flowcited by 39
  24. 2022Disentangling Architecture and Training for Optical Flowcited by 32
  25. 2023TMA: Temporal Motion Aggregation for Event-based Optical Flowcited by 32
  26. 2022Lightweight Event-based Optical Flow Estimation via Iterative Deblurringcited by 31
  27. 2024HDRFlow: Real-Time HDR Video Reconstruction with Large Motionscited by 31
  28. 2023Deep Event Visual Odometrycited by 29
  29. 2023Learning Optical Flow from Event Camera with Rendered Datasetcited by 26
  30. 2023Memory-Efficient Optical Flow via Radius-Distribution Orthogonal Cost Volumecited by 9
  31. 2024Hybrid Cost Volume for Memory-Efficient Optical Flowcited by 6
  32. 2024Leveraging Consistent Spatio-Temporal Correspondence for Robust Visual Odometrycited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026E-WAVE: Event-based Continuous Optical Flow via Warping-Aligned Visual Encodingcited by 0