Simplified Deep Learning Driven Design of Blue Light Grating Couplers
Source abstract
Abstract
ABSTRACT In recent years, visible light photonic integrated circuits have emerged as a promising solution for atomic, quantum, and biosensing applications. However, there remains a lack of comprehensive research on fundamental photonic integrated devices like grating couplers. In this paper, a simplified deep neural network (DNN) incorporating a multi‐head self‐attention mechanism is developed to optimize the design of blue light grating couplers. The simplified DNN model takes only 3 min to complete the training process on a home‐grade computer, facilitating efficient access to the predicted response spectra of the device. The trained DNN model has a custom hybrid error of less than 0.03, with a prediction accuracy as high as 0.95 within a precise range of ±0.02, while the maximum global error is below 0.04. Experimental validation of the DNN‐designed structure achieved a coupling efficiency of ‐5.5 dB at 445 nm with a −3 dB bandwidth of 13 nm, which is in good agreement with the simulation. This work presents a reliable design solution for blue light band grating couplers, laying the foundation for incorporating machine learning techniques in designing visible light photonic devices.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Photonic and Optical Devices · Neural Networks and Reservoir Computing · Photonic Crystals and Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
The MoE Key Laboratory for Information Science of EM Waves College of Future Information Technology Fudan University Shanghai China
Department of Mechanical Engineering, the University of Hong Kong Hong Kong China
Funding
National Natural Science Foundation of China · Shanghai Rising-Star Program
References (44)
Show source-provided references
1. 10.1038/nphoton.2017.9310.1038/nphoton.2017.93 ↗
2. 10.1109/MNANO.2019.289136910.1109/MNANO.2019.2891369 ↗
3. 10.1002/admt.20190113810.1002/admt.201901138 ↗
4. 10.1109/LPT.2021.309831810.1109/LPT.2021.3098318 ↗
5. 10.1002/lpor.20130018310.1002/lpor.201300183 ↗
6. 10.1063/5.005011710.1063/5.0050117 ↗
7. 10.1364/OL.42.00409110.1364/OL.42.004091 ↗
8. 10.1038/s41566‐018‐0236‐y10.1038/s41566‐018‐0236‐y ↗
9. 10.1002/adom.20240313310.1002/adom.202403133 ↗
10. 10.1021/acsphotonics.5b0059910.1021/acsphotonics.5b00599 ↗
11. 10.1038/s41586‐022‐05119‐910.1038/s41586‐022‐05119‐9 ↗
12. 10.1038/s41586‐020‐2823‐610.1038/s41586‐020‐2823‐6 ↗
13. 10.1038/s41586‐020‐2811‐x10.1038/s41586‐020‐2811‐x ↗
14. 10.1364/OPTICA.5.00044310.1364/OPTICA.5.000443 ↗
15. 10.1109/JSTQE.2021.309372110.1109/JSTQE.2021.3093721 ↗
16. 10.1063/1.513168310.1063/1.5131683 ↗
17. 10.1038/s41467‐024‐47105‐x10.1038/s41467‐024‐47105‐x ↗
18. 10.1364/PRJ.51682910.1364/PRJ.516829 ↗
19. 10.1109/JSTQE.2024.347479710.1109/JSTQE.2024.3474797 ↗
20. 10.1364/AOP.50184610.1364/AOP.501846 ↗
21. 10.1364/OE.27.03740010.1364/OE.27.037400 ↗
22. 10.1364/OE.41724510.1364/OE.417245 ↗
23. 10.1364/OME.47987110.1364/OME.479871 ↗
24. 10.1038/s41467‐021‐22046‐x10.1038/s41467‐021‐22046‐x ↗
25. 10.1364/OSAC.2.00115510.1364/OSAC.2.001155 ↗
26. 10.1038/s41566‐020‐0685‐y10.1038/s41566‐020‐0685‐y ↗
27. 10.1002/lpor.20100001410.1002/lpor.201000014 ↗
28. 10.1364/OE.21.02169310.1364/OE.21.021693 ↗
29. 10.1021/acs.nanolett.9b0185710.1021/acs.nanolett.9b01857 ↗
30. 10.1038/s41467‐019‐12698‐110.1038/s41467‐019‐12698‐1 ↗
31. 10.1109/JLT.2017.275567310.1109/JLT.2017.2755673 ↗
32. 10.1109/JSTQE.2018.288548610.1109/JSTQE.2018.2885486 ↗
33. 10.1109/JLT.2021.305747310.1109/JLT.2021.3057473 ↗
34. A.KroghandJ.Hertz “A Simple Weight Decay Can Improve Generalization ”Advances in Neural Information Processing Systems(NIPS1991):4.
35. A. Y.Ng Feature Selection L 1 vs. L 2 Regularization and Rotational Invariance inProceedings of the Twenty‐First International Conference on Machine Learning (ACM Inc.2004):78.
36. A.Vaswani N.Shazeer N.Parmar et al. “Attention Is All You Need ”Advances in Neural Information Processing Systems (NIPS2017).
37. 10.48550/arXiv.1409.047310.48550/arXiv.1409.0473 ↗
38. 10.48550/arXiv.1703.0313010.48550/arXiv.1703.03130 ↗
39. 10.48550/arXiv.2402.1944210.48550/arXiv.2402.19442 ↗
40. 10.1007/s00521‐022‐07366‐310.1007/s00521‐022‐07366‐3 ↗
41. 10.1109/72.27918110.1109/72.279181 ↗
42. 10.48550/arXiv.1710.0594110.48550/arXiv.1710.05941 ↗
43. Neural Architecture Search: A Survey
44. X.Glorot A.Bordes andY.Bengio Deep Sparse Rectifier Neural Networks inProceedingsof the Fourteenth International Conference on Artificial Intelligence and Statistics JMLR Workshop and Conference Proceedings (PMLR2011):315–323.