EAF

<div>The Event-based Autofocus Dataset (EAD):</div><div>We provide all datasets in two formats: MAT file for Matlab and aedat4 for DV. </div><div><br></div><div>We recommend using the aedat4 files since they can be loaded easily using DV. The MAT files are converted from aedar4 for our code. We have included time-synchronized focal positions in the MAT file. </div><div><br></div><div>DV: https://inivation.gitlab.io/dv/dv-docs/docs/getting-started/</div><div><br></div><div>You can directly use it or use the `focus_position.txt`. The format is:</div><div>timestamp_focal_position position</div><div>...</div><div><br></div><div><br></div><div>We also include the png file for visualization. The png file is named by its timestamp. </div><div><br></div><div>`gt_focus_points.txt` is the ground truth focal position after carefully calibration. The format is:</div><div>sequence_name position</div><div>...</div><div><br></div><div>Dataset format:<br></div><div>> gt_focus_points.txt</div><div>> - sequence name</div><div>> images/***.png</div><div>> ***.aedat4</div><div>> ***.mat</div><div>> focus_position.txt</div><div><br></div><div>BibTex</div><div>Please cite our work if you use the data .</div><div><br></div><div>@inproceedings{lin2022autofocus,</div><div> title={Autofocus for Event Cameras},</div><div> author={Shijie, Lin and Yinqiang, Zhang and Lei, Yu and Bin, Zhou and Xiaowei, Luo and Jia, Pan},</div><div> booktitle={The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},</div><div> year={2022},</div><div>}</div><div><br></div>