Datsets used for estimation of event filtering efficiency in the High throughput event filtering: The interpolation-based DIF algorithm hardware architecture paper.
The datasets have been uploaded to the Google drive as the files are too big to be stored directly on GitHub.
https://drive.google.com/drive/folders/1pV0U2GFZuVrmySPI-VtKKd6KYd3xe0-3?usp=sharing
Recorded directory contains datasets recorded with a DVS sensor.
V2E_AGH directory contains datasets generated with the V2E tool.
Recorded_filtered directory contains datasts from Recorded directory, but filtered with NN algorithm.
Recorded_noise directory contains noise recorded with a DVS sensor.
The datasets were used with the noise generator provided in NoiseGeneratorCPP repository.
If you find the resources usefull, please cite the paper.
@article{Kowalczyk2025,
title = {High throughput event filtering: The interpolation-based DIF algorithm hardware architecture},
journal = {Microprocessors and Microsystems},
volume = {117},
pages = {105171},
year = {2025},
issn = {0141-9331},
doi = {https://doi.org/10.1016/j.micpro.2025.105171},
url = {https://www.sciencedirect.com/science/article/pii/S0141933125000390},
author = {Marcin Kowalczyk and Tomasz Kryjak},
keywords = {Dataset, Denoising, DVS, FPGA, High-throughput},
}