Mitigating Noise Effects in Photonic Neural Networks Using Adaptive Quantization
Photonic neural networks offer energy-efficiency but suffer from noise-induced low precision. We propose AQ-PANN, which learns a quantization step size to mitigate noise. Experiments on SVHN show strong performance across bitwidths under different noise levels.
Authors
- Nicola Andriolli (ORCID: https://orcid.org/0000-0001-8322-6854)
- G. Contestabile (ORCID: https://orcid.org/0000-0001-8703-3291)
- I. Roumpos
- P. S. Kincaid
- Luca Valcarenghi (ORCID: https://orcid.org/0000-0002-6695-5032)
- N. Pleros
- Lorenzo De Marinis (ORCID: https://orcid.org/0000-0003-1462-3199)
- Emilio Paolini (ORCID: https://orcid.org/0000-0001-7486-9268)
- M. Moralis-Pegios
Institutions
- University of Pisa (IT)
- Scuola Superiore Sant'Anna (IT)
- Aristotle University of Thessaloniki (GR)
Publication Details
- Published
- 2025-06-29
- DOI
- https://doi.org/10.23919/oecc/psc62146.2025.11110436
- Primary Topic
- Neural Networks and Reservoir Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00