rk EHDRD

EHDRD: A Benchmark for Event-based HDR Imaging

Xiaopeng Li1, Zhaoyuan Zeng1, Cien Fan1, Chen Zhao1, Lei Yu1,
1Wuhan University
SEHDRD BEHDRD

The event-based high dynamic range imaging benchmark EHDRD consists of SEHDRD dataset captured in slow-moving scenes and BEHDRD dataset captured in fast-moving scenes. The generation process of HDR images is inspired by DeepHDRvideo，which be accessed in EHDRD_code.


Introduction of SEHDRD

The SEHDRD Dataset can be downloaded via Baidu Drive, which contains three subdatasets.




  • SEHDRD-S contains HDR-LDR image pairs from Kalantari13, HDM-HDR-2014, and DeepHDRVideo. These datasets contain paired LDR-HDR video sequences which can be leveraged to synthesize events. We utilize the ESIM simulator to synthesize concurrent events.
  • SEHDRD-RSE contains aligned real-world LDR images, HDR image, and real-world event streams, which are captured by FLIR BFS-U3-32S4 camera and SilkyEvCam event camera.
  • SEHDRD-RDV contains aligned real-world LDR images, HDR image, and real-world event streams, which are captured by FLIR BFS-U3-04S2 camera and Davis346 event camera.



Introduction of BEHDRD

The BEHDRD Dataset can be downloaded via Google Drive, which contains three subdatasets:




  • BEHDRD-S contains HDR-LDR image pairs from Kalantari13, HDM-HDR-2014, and DeepHDRVideo. These datasets contain paired LDR-HDR video sequences which can be leveraged to synthesize events. Specifically, the blurry LDR image is obtained by averaging 13 consecutive LDR images, with corresponding HDR images serving as the sharp HDR sequence. We utilize the ESIM simulator \cite{rebecq2018esim} to synthesize concurrent events for each blurry LDR image.
  • BEHDRD-RG contains aligned real-world blurry LDR images, sharp HDR sequence, and real-world event streams images, which are captured by FLIR BFS-U3-32S4 camera, FLIR BFS-U3-04S2 camera, and SilkyEvCam event camera in static scenes.
  • BEHDRD-RGL contains aligned real-world blurry LDR images, sharp LDR images, and real-world event streams images, which are captured by FLIR BFS-U3-32S4 camera, FLIR BFS-U3-04S2 camera, and SilkyEvCam event camera in dynamic scenes.





BibTeX

@misc{xiaopeng2024hdr,
        title={HDR Imaging for Dynamic Scenes with Events}, 
        author={Li Xiaopeng and Zeng Zhaoyuan and Fan Cien and Zhao Chen and Deng Lei and Yu Lei},
        year={2024},
        eprint={2404.03210},
        archivePrefix={arXiv},
        primaryClass={cs.CV}
  }