N-ROD: a Neuromorphic Dataset for Synthetic-to-Real Domain Adaptation
Marco Cannici Politecnico di Milano
Chiara Plizzari Politecnico di Torino
Mirco Planamente Politecnico di Torino, IIT
Marco Ciccone Politecnico di Milano
Andrea Bottino Politecnico di Torino
Barbara Caputo Politecnico di Torino, IIT
Matteo Matteucci Politecnico di Milano
CVPR 2021 Workshop on Event-based Vision
Held in conjunction with the IEEE Conference on Computer Vision and Pattern Recognition 2021

Abstract: Event cameras are novel neuromorphic sensors, which asynchronously capture pixel-level intensity changes in the form of "events". Event simulation from existing RGB datasets is commonly used to overcome the need of large amount of annotated data, which lacks due to the novelty of the event sensors. In this context, the possibility of using event simulation in synthetic scenarios, where data generation is not limited to pre-existing datasets, is to date still unexplored. In this work, we analyze the synth-to-real domain shift in event data, i.e., the gap arising between simulated events obtained from synthetic renderings and those captured with a real camera on real images. To this purpose, we extend to the event modality the popular RGB-D Object Dataset (ROD), which already comes with its synthetic version (SynROD). The resulting extended dataset is the first to enable a synth-to-real analysis on event data. On the proposed Neuromorphic ROD dataset (N-ROD), we show how Domain Adaptation techniques can be used to reduce the synth-to-real gap, thus enabling to take full advantage of automatically generated synthetic data during training. Moreover, through extensive experiments on multi-modal RGB-E data, we show that events can be effectively combined with conventional visual information, encouraging further research in this area.

Neuromorphic ROD

Paper and Bibtex


M. Cannici, C. Plizzari, M. Planamente, M. Ciccone, A. Bottino, B. Caputo, M. Matteucci. N-ROD: a Neuromorphic Dataset for Synthetic-to-Real Domain Adaptation. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2021.

  author    = {Cannici, Marco and Plizzari, Chiara and
               Planamente, Mirco and Ciccone, Marco and
               Bottino, Andrea and Caputo, Barbara and
               Matteucci, Matteo},
  title     = {N-ROD: A Neuromorphic Dataset for
               Synthetic-to-Real Domain Adaptation},
  booktitle = {Proceedings of the IEEE/CVF Conference on
               Computer Vision and Pattern Recognition
               (CVPR) Workshops},
  month     = {June},
  year      = {2021},
  pages     = {1342-1347}

Original Website Template