Welcome to the BlinkVision Benchmark Suite!

BlinkVision is a synthetic benchmark designed primarily for correspondence estimation tasks, including optical flow, scene flow and point tracking. It contains some features:

  1. Two input modalities: RGB frames and events.
  2. Benchmark for correspondence estimation tasks including optical flow, scene flow and point tracking.
  3. Except for the correspondence labels, you can also get access to rich ground truth labels in the training data, including depth, normal, semantic segmentation, camera intrinsic/pose, object pose.
  4. Diverse and complex scenes. Hundreds of daily categories. Photorealistic rendering results. You can preview the dataset in the download page.

Citation

When using this work in an academic context, please cite the following publications:

@inproceedings{blinkflow_iros2023,
    title={BlinkFlow: A Dataset to Push the Limits of Event-based Optical Flow Estimation},
    author={Yijin Li, Zhaoyang Huang, Shuo Chen, Xiaoyu Shi, Hongsheng Li, Hujun Bao, Zhaopeng Cui, Guofeng Zhang},
    booktitle={IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
    month = {October},
    year = {2023},
}
@inproceedings{blinkvision_eccv2024,
    title={BlinkVision: A Benchmark for Optical Flow, Scene Flow and Point Tracking Estimation using RGB Frames and Events},
    author={Yijin Li, Yichen Shen, Zhaoyang Huang, Shuo Chen, Weikang Bian, Xiaoyu Shi, Fu-Yun Wang, Keqiang Sun, Hujun Bao, Zhaopeng Cui, Guofeng Zhang, Hongsheng Li},
    booktitle={European Conference on Computer Vision (ECCV)},
    year={2024}
}

Changelog:

  • 2025.01.12: Release the BlinkVision training data.
  • 2025.01.09: Fix the https license issue.
  • 2023.10.16: The BlinkVision Benchmark Suite goes online, starting with the flow-v1 benchmarks.

License:

This dataset is provided to you under the Creative Commons Attribution-ShareAlike 4.0 International public license (CC BY-SA 4.0). This means that you must attribute the work in the manner specified by the authors and must distribute your contributions under the same license. Using the material for commercial purposes is allowed.