Bolin Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Bolin Li has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Jian Lin
School of Artificial Intelligence Science and Technology, University of Shanghai for Science and Technology 1 , Shanghai
2 shared publications Jinlei Fei
Institute of Photonic Chips, University of Shanghai for Science and Technology 2 , Shanghai
2 shared publications- Min Gu
School of Artificial Intelligence Science and Technology University of Shanghai for Science and Technology Shanghai China
2 shared publications Yinfei Zhu
Institute of Photonic Chips, University of Shanghai for Science and Technology 2 , Shanghai
2 shared publications- Guangrui Luan
School of Artificial Intelligence Science and Technology, University of Shanghai for Science and Technology 1 , Shanghai
1 shared publication - Runshi Zheng
School of Artificial Intelligence Science and Technology, University of Shanghai for Science and Technology 1 , Shanghai
1 shared publication
Latest Publications
- 1 February 2025
- Published by AIP Publishing in Applied Physics Letters
- Vol. 126 (8)
Improved spectral filtering of broadband diffractive neural network by loss function engineering
We engineer the loss function by removing the conventional physics-based energy constraint during the training of broadband diffractive neural networks (DNNs) to enhance their spectral filtering capabilities of supercontinuum light. Simulations show that compared to DNNs trained with conventional loss function, the suppression of out-of-band spectral intensities can be improved by three orders of magnitude, resulting in an extinction coefficient of 10−6. Additionally, the spectral resolution can be enhanced by over 50% with a 6.6% improvement of energy efficiency. These findings are corroborated by experiments conducted with a two-layer DNN. The proposed method holds promise for enhancing the performance of broadband DNNs across various applications, including spectral reconstruction, spectrum classification, and color image processing, among others.
- 1 February 2025
- Published by AIP Publishing in Applied Physics Letters
- Vol. 126 (8)
Improved spectral filtering of broadband diffractive neural network by loss function engineering
We engineer the loss function by removing the conventional physics-based energy constraint during the training of broadband diffractive neural networks (DNNs) to enhance their spectral filtering capabilities of supercontinuum light. Simulations show that compared to DNNs trained with conventional loss function, the suppression of out-of-band spectral intensities can be improved by three orders of magnitude, resulting in an extinction coefficient of 10−6. Additionally, the spectral resolution can be enhanced by over 50% with a 6.6% improvement of energy efficiency. These findings are corroborated by experiments conducted with a two-layer DNN. The proposed method holds promise for enhancing the performance of broadband DNNs across various applications, including spectral reconstruction, spectrum classification, and color image processing, among others.
- Open Access
- 1 January 2025
- Published by AIP Publishing in APL Photonics
- Vol. 10 (1)
Multi-functional broadband diffractive neural network with a single spatial light modulator
Diffractive neural networks (DNNs) are emerging as a novel optical computing architecture that combines wave optics with deep-learning methods for high-speed parallel information processing. Herein, we report a reflection type, multi-functional, broadband DNN design. It consists of two phase-modulation layers based on a single spatial light modulator and a mirror facing it. The power efficiency of this design is more than 16 times higher than that of the cascaded structure utilizing beam splitters. It can function either as a two-layer DNN or a one-layer DNN with the other serving as an information input layer. Single- and dual-wavelength filtering and focusing, as well as spatial wavelength demultiplexing of supercontinuum, are experimentally demonstrated using the two-layer DNN, whereas the one-layer DNN is experimentally demonstrated by the classification of hand-written digits, which are input by the first layer via holographic imaging. The designed DNN could operate independently or be readily integrated with other optical systems and may find applications in spectroscopy, microscopy, and information technology.
- 1 January 2025
- Published by AIP Publishing in APL Photonics
- Vol. 10 (1)
Open AccessMulti-functional broadband diffractive neural network with a single spatial light modulator
Diffractive neural networks (DNNs) are emerging as a novel optical computing architecture that combines wave optics with deep-learning methods for high-speed parallel information processing. Herein, we report a reflection type, multi-functional, broadband DNN design. It consists of two phase-modulation layers based on a single spatial light modulator and a mirror facing it. The power efficiency of this design is more than 16 times higher than that of the cascaded structure utilizing beam splitters. It can function either as a two-layer DNN or a one-layer DNN with the other serving as an information input layer. Single- and dual-wavelength filtering and focusing, as well as spatial wavelength demultiplexing of supercontinuum, are experimentally demonstrated using the two-layer DNN, whereas the one-layer DNN is experimentally demonstrated by the classification of hand-written digits, which are input by the first layer via holographic imaging. The designed DNN could operate independently or be readily integrated with other optical systems and may find applications in spectroscopy, microscopy, and information technology.