{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T22:50:57Z","timestamp":1780613457505,"version":"3.54.1"},"reference-count":36,"publisher":"Springer Science and Business Media LLC","issue":"5","license":[{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T00:00:00Z","timestamp":1772841600000},"content-version":"vor","delay-in-days":6,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100007176","name":"Scuola Superiore Sant'Anna","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100007176","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Neural Comput &amp; Applic"],"published-print":{"date-parts":[[2026,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Photonic neural networks have emerged as a promising solution to overcome limitations of traditional hardware for neuromorphic computations, offering advantages in bandwidth, latency, and power efficiency. However, their performance is constrained by the limited precision of analog photonic computing, which is affected by inherent noise sources such as thermal and shot noise, and distortions. These effects degrade the photonic neural network performance, reducing the bit resolution achievable in photonic hardware typically to 2\u20134 bits. Traditional quantization strategies fail to account for these noise contributions, resulting in a substantial accuracy loss during inference. This paper introduces an adaptive quantization method called Adaptive-Quantization Photonic-Aware Neural Network (AQ-PANN) to address the challenges posed by different noise sources in analog photonic hardware. The proposed method uses a learnable step size quantization scheme to achieve high accuracy and stability under varying noise levels, introducing a scheme that unifies quantization step adaptation with noise injection exactly where photonic distortions arise. This design incurs only a minor training-time overhead, as it involves learning a small number of per-layer quantization step sizes and does not affect inference. Experimental evaluations on three commonly used test datasets (MNIST, SVHN, and CIFAR-10) with different bit resolutions show the robustness of AQ-PANN. On MNIST, an accuracy drop of only 2% was observed from low to high noise levels in a 4-bit configuration, while traditional DoReFa quantization suffered a 29% drop. For the SVHN dataset, AQ-PANN obtained a mean accuracy of 92% under high noise with 4-bit quantization, outperforming DoReFa by over 45%. On CIFAR-10, AQ-PANN maintained close to 60% accuracy under high noise in the 4-bit configuration, whereas DoReFa and PACT both collapsed below 40%. These results highlight the effectiveness of AQ-PANN in sustaining model performance across different noise intensities, enabling practical photonic neural network deployment.<\/jats:p>","DOI":"10.1007\/s00521-026-11984-6","type":"journal-article","created":{"date-parts":[[2026,3,7]],"date-time":"2026-03-07T06:54:33Z","timestamp":1772866473000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Noise-resilient photonic neural networks through adaptive quantization"],"prefix":"10.1007","volume":"38","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7486-9268","authenticated-orcid":false,"given":"Emilio","family":"Paolini","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lorenzo","family":"De Marinis","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter Seigo","family":"Kincaid","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Luca","family":"Valcarenghi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Giampiero","family":"Contestabile","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ioannis","family":"Roumpos","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Miltiadis","family":"Moralis-Pegios","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nikos","family":"Pleros","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nicola","family":"Andriolli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,3,7]]},"reference":[{"issue":"1","key":"11984_CR1","doi-asserted-by":"publisher","first-page":"7068349","DOI":"10.1155\/2018\/7068349","volume":"2018","author":"A Voulodimos","year":"2018","unstructured":"Voulodimos A, Doulamis N, Doulamis A, Protopapadakis E (2018) Deep learning for computer vision: a brief review. Comput Intell Neurosci 2018(1):7068349. https:\/\/doi.org\/10.1155\/2018\/7068349","journal-title":"Comput Intell Neurosci"},{"issue":"1","key":"11984_CR2","doi-asserted-by":"publisher","first-page":"538","DOI":"10.1109\/COMST.2022.3233793","volume":"25","author":"K He","year":"2023","unstructured":"He K, Kim DD, Asghar MR (2023) Adversarial machine learning for network intrusion detection systems: a comprehensive survey. IEEE Commun Surv Tutor 25(1):538\u2013566. https:\/\/doi.org\/10.1109\/COMST.2022.3233793","journal-title":"IEEE Commun Surv Tutor"},{"key":"11984_CR3","doi-asserted-by":"publisher","first-page":"1155","DOI":"10.1109\/OJCOMS.2024.3363082","volume":"5","author":"E Paolini","year":"2024","unstructured":"Paolini E, Valcarenghi L, Maggiani L, Andriolli N (2024) Real-time network packet classification exploiting computer vision architectures. IEEE Open J Commun Soc 5:1155\u20131166. https:\/\/doi.org\/10.1109\/OJCOMS.2024.3363082","journal-title":"IEEE Open J Commun Soc"},{"key":"11984_CR4","unstructured":"Wang N, Choi J, Brand D, Chen C-Y, Gopalakrishnan K (2018) Training deep neural networks with 8-bit floating point numbers. In: Proceedings of the 32nd International Conference on Neural Information Processing Systems. NIPS\u201918, pp. 7686\u20137695. Curran Associates Inc., Red Hook, NY, USA"},{"key":"11984_CR5","doi-asserted-by":"publisher","unstructured":"Paolini E, De Marinis L, Cococcioni M, Valcarenghi L, Maggiani L, Andriolli N (2022) Photonic-aware neural network: a fixed-point emulation of photonic hardware. In: 2022 27th OptoElectronics and Communications Conference (OECC) and 2022 International Conference on Photonics in Switching and Computing (PSC), pp. 01\u201303. https:\/\/doi.org\/10.23919\/OECC\/PSC53152.2022.9850019","DOI":"10.23919\/OECC\/PSC53152.2022.9850019"},{"issue":"4","key":"11984_CR6","doi-asserted-by":"publisher","first-page":"485","DOI":"10.1109\/JPROC.2020.2976475","volume":"108","author":"L Deng","year":"2020","unstructured":"Deng L, Li G, Han S, Shi L, Xie Y (2020) Model compression and hardware acceleration for neural networks: a comprehensive survey. Proc IEEE 108(4):485\u2013532. https:\/\/doi.org\/10.1109\/JPROC.2020.2976475","journal-title":"Proc IEEE"},{"issue":"2","key":"11984_CR7","doi-asserted-by":"publisher","first-page":"2312825","DOI":"10.1002\/adma.202312825","volume":"37","author":"R Li","year":"2024","unstructured":"Li R, Gong Y, Huang H, Zhou Y, Mao S, Wei Z, Zhang Z (2024) Photonics for neuromorphic computing: fundamentals, devices, and opportunities. Adv Mater 37(2):2312825","journal-title":"Adv Mater"},{"issue":"18","key":"11984_CR8","doi-asserted-by":"publisher","first-page":"15589","DOI":"10.1007\/s00521-022-07243-z","volume":"34","author":"E Paolini","year":"2022","unstructured":"Paolini E, De Marinis L, Cococcioni M, Valcarenghi L, Maggiani L, Andriolli N (2022) Photonic-aware neural networks. Neural Comput Appl 34(18):15589\u201315601","journal-title":"Neural Comput Appl"},{"issue":"1","key":"11984_CR9","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/JSTQE.2019.2941485","volume":"26","author":"MA Nahmias","year":"2019","unstructured":"Nahmias MA, De Lima TF, Tait AN, Peng H-T, Shastri BJ, Prucnal PR (2019) Photonic multiply-accumulate operations for neural networks. IEEE J Sel Top Quantum Electron 26(1):1\u201318","journal-title":"IEEE J Sel Top Quantum Electron"},{"issue":"12","key":"11984_CR10","doi-asserted-by":"publisher","first-page":"1335","DOI":"10.1038\/s41566-024-01567-z","volume":"18","author":"S Bandyopadhyay","year":"2024","unstructured":"Bandyopadhyay S, Sludds A, Krastanov S, Hamerly R, Harris N, Bunandar D, Streshinsky M, Hochberg M, Englund D (2024) Single-chip photonic deep neural network with forward-only training. Nat Photonics 18(12):1335\u20131343. https:\/\/doi.org\/10.1038\/s41566-024-01567-z","journal-title":"Nat Photonics"},{"issue":"1","key":"11984_CR11","doi-asserted-by":"publisher","first-page":"9081","DOI":"10.1038\/s41467-024-53261-x","volume":"15","author":"Z Lin","year":"2024","unstructured":"Lin Z, Shastri BJ, Yu S, Song J, Zhu Y, Safarnejadian A, Cai W, Lin Y, Ke W, Hammood M et al (2024) 120 GOPS Photonic tensor core in thin-film lithium niobate for inference and in situ training. Nat Commun 15(1):9081","journal-title":"Nat Commun"},{"key":"11984_CR12","doi-asserted-by":"crossref","unstructured":"Hu Y, Song Y, Zhu X, Guo X, Lu S, Zhang Q, He L, Franken C, Powell K, Warner H et al (2024) Integrated lithium niobate photonic computing circuit based on efficient and high-speed electro-optic conversion. arXiv preprint arXiv:2411.02734","DOI":"10.1038\/s41467-025-62635-8"},{"issue":"5","key":"11984_CR13","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/JQE.2022.3177793","volume":"58","author":"L De Marinis","year":"2022","unstructured":"De Marinis L, Catania A, Castoldi P, Contestabile G, Bruschi P, Piotto M, Andriolli N (2022) A codesigned integrated photonic electronic neuron. IEEE J Quantum Electron 58(5):1\u201310. https:\/\/doi.org\/10.1109\/JQE.2022.3177793","journal-title":"IEEE J Quantum Electron"},{"issue":"20","key":"11984_CR14","doi-asserted-by":"publisher","first-page":"34264","DOI":"10.1364\/OE.532306","volume":"32","author":"I Roumpos","year":"2024","unstructured":"Roumpos I, Marinis LD, Kovaios S, Kincaid PS, Paolini E, Tsakyridis A, Moralis-Pegios M, Berciano M, Ferraro F, Bode D, Srinivasan SA, Pantouvaki M, Andriolli N, Contestabile G, Pleros N, Vyrsokinos K (2024) Silicon integrated photonic-electronic neuron for noise-resilient deep learning. Opt Express 32(20):34264\u201334274. https:\/\/doi.org\/10.1364\/OE.532306","journal-title":"Opt Express"},{"key":"11984_CR15","doi-asserted-by":"publisher","unstructured":"Paolini E, De Marinis L, Contestabile G, Gupta S, Maggiani L, Andriolli N (2023) Validation of Photonic Neural Networks in Health Scenarios. In: 2023 International Conference on Photonics in Switching and Computing (PSC). https:\/\/doi.org\/10.1109\/PSC57974.2023.10297132","DOI":"10.1109\/PSC57974.2023.10297132"},{"key":"11984_CR16","doi-asserted-by":"publisher","unstructured":"Paolini E, Civerchia F, De Marinis L, Valcarenghi L, Maggiani L, Andriolli N (2022) Photonic-aware Neural Networks for Packet Classification in Beyond 5G Networks. In: 2022 13th International Conference on Network of the Future (NoF). https:\/\/doi.org\/10.1109\/NoF55974.2022.9942486","DOI":"10.1109\/NoF55974.2022.9942486"},{"issue":"1","key":"11984_CR17","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/JSTQE.2019.2945548","volume":"26","author":"B Shi","year":"2020","unstructured":"Shi B, Calabretta N, Stabile R (2020) Deep neural network through an InP SOA-based photonic integrated cross-connect. IEEE J Sel Top Quantum Electron 26(1):1\u201311. https:\/\/doi.org\/10.1109\/JSTQE.2019.2945548","journal-title":"IEEE J Sel Top Quantum Electron"},{"issue":"1","key":"11984_CR18","doi-asserted-by":"publisher","first-page":"140","DOI":"10.1109\/TETCI.2022.3182765","volume":"7","author":"M Kirtas","year":"2023","unstructured":"Kirtas M, Passalis N, Mourgias-Alexandris G, Dabos G, Pleros N, Tefas A (2023) Robust architecture-agnostic and noise resilient training of photonic deep learning models. IEEE Trans Emerg Top Comput Intell 7(1):140\u2013149. https:\/\/doi.org\/10.1109\/TETCI.2022.3182765","journal-title":"IEEE Trans Emerg Top Comput Intell"},{"issue":"1","key":"11984_CR19","doi-asserted-by":"publisher","DOI":"10.1063\/5.0169810","volume":"9","author":"A Tsakyridis","year":"2024","unstructured":"Tsakyridis A, Moralis-Pegios M, Giamougiannis G, Kirtas M, Passalis N, Tefas A, Pleros N (2024) Photonic neural networks and optics-informed deep learning fundamentals. APL Photonics 9(1):011102. https:\/\/doi.org\/10.1063\/5.0169810","journal-title":"APL Photonics"},{"key":"11984_CR20","doi-asserted-by":"publisher","first-page":"561","DOI":"10.1016\/j.neunet.2022.09.015","volume":"155","author":"M Kirtas","year":"2022","unstructured":"Kirtas M, Oikonomou A, Passalis N, Mourgias-Alexandris G, Moralis-Pegios M, Pleros N, Tefas A (2022) Quantization-aware training for low precision photonic neural networks. Neural Netw 155:561\u2013573","journal-title":"Neural Netw"},{"key":"11984_CR21","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-48655-5","volume-title":"Solitonic neural networks","author":"A Bile","year":"2024","unstructured":"Bile A (2024) Solitonic neural networks. Springer Nature, Cham"},{"issue":"11","key":"11984_CR22","doi-asserted-by":"publisher","first-page":"5585","DOI":"10.3390\/app12115585","volume":"12","author":"A Bile","year":"2022","unstructured":"Bile A, Tari H, Fazio E (2022) Episodic memory and information recognition using solitonic neural networks based on photorefractive plasticity. Appl Sci 12(11):5585","journal-title":"Appl Sci"},{"issue":"11","key":"11984_CR23","doi-asserted-by":"publisher","first-page":"1756","DOI":"10.3390\/nano13111756","volume":"13","author":"T Li","year":"2023","unstructured":"Li T, Li Y, Wang Y, Liu Y, Liu Y, Wang Z, Miao R, Han D, Hui Z, Li W (2023) Neuromorphic photonics based on phase change materials. Nanomaterials 13(11):1756","journal-title":"Nanomaterials"},{"issue":"1","key":"11984_CR24","doi-asserted-by":"publisher","first-page":"4874","DOI":"10.1038\/s41598-022-08703-1","volume":"12","author":"J Robertson","year":"2022","unstructured":"Robertson J, Kirkland P, Alanis JA, Hejda M, Bueno J, Di Caterina G, Hurtado A (2022) Ultrafast neuromorphic photonic image processing with a vcsel neuron. Sci Rep 12(1):4874","journal-title":"Sci Rep"},{"key":"11984_CR25","doi-asserted-by":"crossref","unstructured":"Owen-Newns D, Robertson J, Hejda M, Hurtado A (2022) Ghz rate neuromorphic photonic spiking neural network with a single vertical-cavity surface-emitting laser (vcsel). IEEE J Sel Top Quantum Electron 29(2: Optical Computing):1\u201310","DOI":"10.1109\/JSTQE.2022.3205716"},{"key":"11984_CR26","unstructured":"Esser SK, McKinstry JL, Bablani D, Appuswamy R, Modha DS (2019) Learned step size quantization. arXiv preprint arXiv:1902.08153"},{"issue":"1","key":"11984_CR27","doi-asserted-by":"publisher","first-page":"5572","DOI":"10.1038\/s41467-022-33259-z","volume":"13","author":"G Mourgias-Alexandris","year":"2022","unstructured":"Mourgias-Alexandris G, Moralis-Pegios M, Tsakyridis A, Simos S, Dabos G, Totovic A, Passalis N, Kirtas M, Rutirawut T, Gardes F et al (2022) Noise-resilient and high-speed deep learning with coherent silicon photonics. Nat Commun 13(1):5572","journal-title":"Nat Commun"},{"key":"11984_CR28","doi-asserted-by":"publisher","unstructured":"Paolini E, De Marinis L, Valcarenghi L, Maggiani L, Andriolli N (2024) Activation stretching for tackling noise in photonic aware neural networks. In: 2024 Optical Fiber Communications Conference and Exhibition (OFC), pp. 2\u201313. https:\/\/doi.org\/10.1364\/OFC.2024.Th2A.13","DOI":"10.1364\/OFC.2024.Th2A.13"},{"key":"11984_CR29","doi-asserted-by":"publisher","unstructured":"Passalis N, Kirtas M, Mourgias-Alexandris G, Dabos G, Pleros N, Tefas A (2021) Training noise-resilient recurrent photonic networks for financial time series analysis. In: 2020 28th European Signal Processing Conference (EUSIPCO), pp. 1556\u20131560. https:\/\/doi.org\/10.23919\/Eusipco47968.2020.9287649","DOI":"10.23919\/Eusipco47968.2020.9287649"},{"key":"11984_CR30","unstructured":"Zhou S, Wu Y, Ni Z, Zhou X, Wen H, Zou Y (2016) Dorefa-net: Training low bitwidth convolutional neural networks with low bitwidth gradients. arXiv preprint arXiv:1606.06160"},{"key":"11984_CR31","unstructured":"Choi J, Wang Z, Venkataramani S, Chuang PI-J, Srinivasan V, Gopalakrishnan K (2018) Pact: Parameterized clipping activation for quantized neural networks. arXiv preprint arXiv:1805.06085"},{"issue":"2","key":"11984_CR32","doi-asserted-by":"publisher","first-page":"329","DOI":"10.1515\/nanoph-2021-0521","volume":"11","author":"Y Tian","year":"2022","unstructured":"Tian Y, Zhao Y, Liu S, Li Q, Wang W, Feng J, Guo J (2022) Scalable and compact photonic neural chip with low learning-capability-loss. Nanophotonics 11(2):329\u2013344","journal-title":"Nanophotonics"},{"key":"11984_CR33","unstructured":"Yin P, Lyu J, Zhang S, Osher S, Qi Y, Xin J (2019) Understanding straight-through estimator in training activation quantized neural nets. arXiv preprint arXiv:1903.05662"},{"issue":"11","key":"11984_CR34","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","volume":"86","author":"Y LeCun","year":"2002","unstructured":"LeCun Y, Bottou L, Bengio Y, Haffner P (2002) Gradient-based learning applied to document recognition. Proc IEEE 86(11):2278\u20132324","journal-title":"Proc IEEE"},{"key":"11984_CR35","unstructured":"Netzer Y, Wang T, Coates A, Bissacco A, Wu B, Ng AY et al (2011) Reading digits in natural images with unsupervised feature learning. In: NIPS Workshop on Deep Learning and Unsupervised Feature Learning, vol. 2011, p. 7. Granada"},{"key":"11984_CR36","unstructured":"Krizhevsky A, Nair V, Hinton G CIFAR-10 (Canadian Institute for Advanced Research)"}],"container-title":["Neural Computing and Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00521-026-11984-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00521-026-11984-6","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00521-026-11984-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,6]],"date-time":"2026-04-06T06:49:38Z","timestamp":1775458178000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00521-026-11984-6"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3]]},"references-count":36,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2026,3]]}},"alternative-id":["11984"],"URL":"https:\/\/doi.org\/10.1007\/s00521-026-11984-6","relation":{},"ISSN":["0941-0643","1433-3058"],"issn-type":[{"value":"0941-0643","type":"print"},{"value":"1433-3058","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3]]},"assertion":[{"value":"17 July 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"3 February 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"7 March 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare that they have no Conflict of interest.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}}],"article-number":"137"}}