Skip to content

Latest commit

 

History

History
61 lines (41 loc) · 3.03 KB

README.md

File metadata and controls

61 lines (41 loc) · 3.03 KB
Raw

DA4Event

Introduction

This repository contains the implementation of the model presented in the following paper:

DA4Event: towards bridging the Sim-to-Real Gap for Event Cameras using Domain Adaptation, Mirco Planamente, Chiara Plizzari, Marco Cannici, Marco Ciccone, Francesco Strada, Andrea Bottino, Matteo Matteucci, Barbara Caputo ArXiv

Alt Text

Prerequisites

Code

Coming soon

Datasets

Here are described the procedure to reproduce our datasets. Please contact us if you want to have direct access to them.

N-Caltech101

Alt Text

Real N-Caltech: the Neuromorphic Caltech101 (N-Caltech101) is an event-based conversion of the popular image dataset Caltech101.

Sim N-Caltech: we used the event simulator ESIM to extract the simulated event streams from Caltech101 RGB images. The conversion from still images to video sequences has been done following Video-to-Events.

RGB-D Object Dataset (ROD)

Alt Text

RGB-E Real ROD: the RGB-D Object Dataset (ROD) is one of the most common benchmarks for object recognition in robotics. We extended it to the event modality by converting the crops provided in RGB-crops to events, following the procedure in Video-to-Events and using ESIM simulator.

RGB-E Syn ROD: a synthetic version of ROD (synROD) has been recently proposed in synROD. We converted the crops provided in syn-RGB-crops to events with the same procedure described above, obtaining a synthetic event version.

Results

N-Caltech101

Alt Text

ROD

Alt Text

Citation

Please cite the following paper if you use this code for your researches:

 @misc{planamente2021da4event,
      title={DA4Event: towards bridging the Sim-to-Real Gap for Event Cameras using Domain Adaptation}, 
      author={Mirco Planamente and Chiara Plizzari and Marco Cannici and Marco Ciccone and Francesco Strada and Andrea Bottino and Matteo Matteucci and Barbara Caputo},
      year={2021},
      eprint={2103.12768},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Contact 📌

If you have any question, do not hesitate to contact us: