Power-over-fiber (PoF) systems suffer from energy waste and instability under dynamic conditions due to fixed laser output and time-varying load demands. This study presents a closed-loop feedback control mechanism based on capacitor-voltage monitoring, which effectively filters transient disturbances while accurately reflecting system energy balance. Employing an 808 nm laser with 12 W maximum output, the system uses a GaAs-based photovoltaic power converter (PPC) with high measured MPP efficiency under our test conditions and a Proportional-Integral-Derivative (PID) algorithm to achieve rapid response, automatic compensation for fiber losses up to 1500 m, and 17.9% end-to-end efficiency over 500 m fiber, while ensuring long-term reliability with stable operation exceeding 3000 s. A practical WiFi access scenario with live video streaming is further demonstrated, validating stable closed-loop powering under traffic-driven load dynamics.
{"title":"Closed-Loop Power-Over-Fiber System With Real-Time Feedback Control for Dynamic Load Adaptation","authors":"Guangxin Li;Zhiguo Zhang;Rui Zhou;Xueliang Gu;Zhehao Yan;Tong Zhai;Yang Zhao","doi":"10.1109/JPHOT.2026.3665995","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3665995","url":null,"abstract":"Power-over-fiber (PoF) systems suffer from energy waste and instability under dynamic conditions due to fixed laser output and time-varying load demands. This study presents a closed-loop feedback control mechanism based on capacitor-voltage monitoring, which effectively filters transient disturbances while accurately reflecting system energy balance. Employing an 808 nm laser with 12 W maximum output, the system uses a GaAs-based photovoltaic power converter (PPC) with high measured MPP efficiency under our test conditions and a Proportional-Integral-Derivative (PID) algorithm to achieve rapid response, automatic compensation for fiber losses up to 1500 m, and 17.9% end-to-end efficiency over 500 m fiber, while ensuring long-term reliability with stable operation exceeding 3000 s. A practical WiFi access scenario with live video streaming is further demonstrated, validating stable closed-loop powering under traffic-driven load dynamics.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-13"},"PeriodicalIF":2.4,"publicationDate":"2026-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11425014","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147440587","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-03DOI: 10.1109/JPHOT.2026.3669963
Yang Wang;Wei Chen;Xiaobei Zhang;Qi Zhang;Tingyun Wang
We demonstrate a double-clad anti-resonant hollow-core fiber (DC-ARF) with an offset inner cladding structure for effective mode tailoring. The DC-ARF consists of six thin-walled, azimuthally offset inner tubes enclosed within a thicker outer cladding tube, which is further connected to silica jacket through a single node. Asymmetric offset arrangement of inner cladding tubes generates two enlarged gaps based on mode field distribution of LP11-like mode, ensuring high-order mode (HOM) suppression. Numerical simulations show that light guidance is achieved through leakage from the larger gaps followed by anti-resonant reflection back into the core. And wavelength disparity between fundamental mode and HOMs provides inherent mechanism for preferential HOM suppression. Through structural optimization, significantly improved mode purity is achieved. The fabricated DC-ARF features two large azimuthal gaps of 70° and operates at a wavelength of 1.06 μm. Experimental results confirm that only the fundamental mode remains when the fiber length exceeds 0.5 m. Furthermore, a 1-m-long DC-ARF improves the input beam quality by approximately 40%, demonstrating its mode tailoring capability. Our work offers valuable guidance for mode tailoring, and the proposed DC-ARF has great potential in single-mode laser transmission, biomedical systems, and optical communications.
{"title":"Double-Clad Anti-Resonant Hollow-Core Fiber With Inner Cladding Tube Offset for Mode Tailoring","authors":"Yang Wang;Wei Chen;Xiaobei Zhang;Qi Zhang;Tingyun Wang","doi":"10.1109/JPHOT.2026.3669963","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3669963","url":null,"abstract":"We demonstrate a double-clad anti-resonant hollow-core fiber (DC-ARF) with an offset inner cladding structure for effective mode tailoring. The DC-ARF consists of six thin-walled, azimuthally offset inner tubes enclosed within a thicker outer cladding tube, which is further connected to silica jacket through a single node. Asymmetric offset arrangement of inner cladding tubes generates two enlarged gaps based on mode field distribution of LP<sub>11</sub>-like mode, ensuring high-order mode (HOM) suppression. Numerical simulations show that light guidance is achieved through leakage from the larger gaps followed by anti-resonant reflection back into the core. And wavelength disparity between fundamental mode and HOMs provides inherent mechanism for preferential HOM suppression. Through structural optimization, significantly improved mode purity is achieved. The fabricated DC-ARF features two large azimuthal gaps of 70° and operates at a wavelength of 1.06 μm. Experimental results confirm that only the fundamental mode remains when the fiber length exceeds 0.5 m. Furthermore, a 1-m-long DC-ARF improves the input beam quality by approximately 40%, demonstrating its mode tailoring capability. Our work offers valuable guidance for mode tailoring, and the proposed DC-ARF has great potential in single-mode laser transmission, biomedical systems, and optical communications.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-7"},"PeriodicalIF":2.4,"publicationDate":"2026-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11419729","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147440498","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Diffractive deep neural network (D2NN), a type of ONN that processes images using light propagation through free space, has been shown to be capable of performing various image processing tasks, yet it still relies partially on electrical signals. In this report, we achieved completely all-optical and continuous real-time processing of two-dimensional visible information directly using light from real objects without converting the input information into any electrical signals. Firstly, two image processing tasks, classification (MNIST) and detection of industrial parts, were performed as a verification of the capability of our novel D2NN optical information processor. Then, high-speed operation was investigated by two further tasks, classification of glass beads and detection of shapes printed on transparent sheets, in which the visible light from the sample was introduced to and processed by the processor without conversion to electrical signal. The results show that our implementation of a D2NN processor is capable of image processing on the nano-second order from the appearance of the sample in the observation area.
衍射深度神经网络（Diffractive deep neural network, D2NN）是一种利用光在自由空间传播来处理图像的神经网络，已被证明能够执行各种图像处理任务，但它仍然部分依赖于电信号。在本报告中，我们在不将输入信息转换为任何电信号的情况下，直接利用真实物体的光，实现了二维可见信息的完全全光连续实时处理。首先，进行了两项图像处理任务，分类（MNIST）和工业零件检测，以验证我们的新型D2NN光信息处理器的能力。然后，通过两项进一步的任务，即玻璃珠的分类和打印在透明片上的形状检测来研究高速运行，其中来自样品的可见光被引入处理器并进行处理，而不转换为电信号。结果表明，我们实现的D2NN处理器能够从观察区域的样品外观进行纳秒级的图像处理。
{"title":"Continuous Real-Time Image Processing by Variable Optical Processor","authors":"Shun Miura;Mamoru Otake;Takahiro Nambara;Hiroyuki Kusaka;Yuichiro Kunai;Masahiro Kashiwagi","doi":"10.1109/JPHOT.2026.3669062","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3669062","url":null,"abstract":"Diffractive deep neural network (D<sup>2</sup>NN), a type of ONN that processes images using light propagation through free space, has been shown to be capable of performing various image processing tasks, yet it still relies partially on electrical signals. In this report, we achieved completely all-optical and continuous real-time processing of two-dimensional visible information directly using light from real objects without converting the input information into any electrical signals. Firstly, two image processing tasks, classification (MNIST) and detection of industrial parts, were performed as a verification of the capability of our novel D<sup>2</sup>NN optical information processor. Then, high-speed operation was investigated by two further tasks, classification of glass beads and detection of shapes printed on transparent sheets, in which the visible light from the sample was introduced to and processed by the processor without conversion to electrical signal. The results show that our implementation of a D<sup>2</sup>NN processor is capable of image processing on the nano-second order from the appearance of the sample in the observation area.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-8"},"PeriodicalIF":2.4,"publicationDate":"2026-03-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11417837","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147440607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-25DOI: 10.1109/JPHOT.2026.3667908
Ziheng Zhang;Zicai Cao;Mengfan Cheng;Zhijun Yan;Qi Yang;Ming Tang;Deming Liu;Lei Deng
As the per-channel data rate in Nyquist WDM systems increases, widely linear (WL) MIMO equalizers are required to compensate transceiver linear impairments. However, the crosstalk, filtering, and mixing effects arising from the interaction between frequency offset (FO) and transceiver impairments significantly degrade the FO tolerance of the WL-MIMO equalizer, limiting its applicability in Nyquist WDM systems. In this paper, we introduce an FO pre-compensation 8 × 2 WL-MIMO equalizer. By integrating a specially designed single-branch FO compensation structure with an 8 × 2 WL-MIMO equalizer based on the digital phase-locked loop least mean square algorithm, the proposed method achieves FO-insensitive transceiver impairment compensation with nearly unchanged computational complexity. Simulations of a Nyquist WDM system show that the proposed scheme achieves an FO tolerance exceeding 5 GHz. Moreover, a 32 GBaud dual-polarization 64 quadrature amplitude modulation signals transmission experiment over 75 km standard single-mode fiber further confirms that, while maintaining stable transceiver impairment compensation, the proposed scheme significantly extends the FO tolerance from 200 MHz to beyond 2 GHz compared with the conventional 8 × 2 WL-MIMO equalizer.
{"title":"Frequency-Offset-Insensitive Equalizer for Coherent Transceivers Linear Impairment Compensation in Nyquist WDM System","authors":"Ziheng Zhang;Zicai Cao;Mengfan Cheng;Zhijun Yan;Qi Yang;Ming Tang;Deming Liu;Lei Deng","doi":"10.1109/JPHOT.2026.3667908","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3667908","url":null,"abstract":"As the per-channel data rate in Nyquist WDM systems increases, widely linear (WL) MIMO equalizers are required to compensate transceiver linear impairments. However, the crosstalk, filtering, and mixing effects arising from the interaction between frequency offset (FO) and transceiver impairments significantly degrade the FO tolerance of the WL-MIMO equalizer, limiting its applicability in Nyquist WDM systems. In this paper, we introduce an FO pre-compensation 8 × 2 WL-MIMO equalizer. By integrating a specially designed single-branch FO compensation structure with an 8 × 2 WL-MIMO equalizer based on the digital phase-locked loop least mean square algorithm, the proposed method achieves FO-insensitive transceiver impairment compensation with nearly unchanged computational complexity. Simulations of a Nyquist WDM system show that the proposed scheme achieves an FO tolerance exceeding 5 GHz. Moreover, a 32 GBaud dual-polarization 64 quadrature amplitude modulation signals transmission experiment over 75 km standard single-mode fiber further confirms that, while maintaining stable transceiver impairment compensation, the proposed scheme significantly extends the FO tolerance from 200 MHz to beyond 2 GHz compared with the conventional 8 × 2 WL-MIMO equalizer.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-8"},"PeriodicalIF":2.4,"publicationDate":"2026-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11410515","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147362445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-17DOI: 10.1109/JPHOT.2026.3665803
Fu-Ming Ya;Xue-Jun Zhang;Pin-Pin Qin
In the free-space optical communication (FSO) domain, orbital angular momentum (OAM) beams are vulnerable to disturbances from composite channels, including atmospheric turbulence and diffuse media, creating significant challenges for accurate mode identification. The present study investigates the transmission characteristics and mode recognition issues of a novel Bessel-Bessel-Gaussian (BBG) vortex beam after traversing a composite channel comprising atmospheric turbulence and diffuse media. Integrating the Fresnel diffraction model with the von Kalman turbulence model enables the numerical simulation of the optical field evolution of BBG beams under varying turbulence intensities and transmission distances. Simulation results in constructing a dataset comprising 9,600 distorted spot images. The evaluation of the recognition performance of multiple deep learning models, including traditional convolutional neural networks (VGG16, ResNet50) and advanced Transformer architectures (Swin-Transformer), is facilitated by this dataset. The simulation results demonstrate that BBG beams exhibit superior shape retention capabilities in complex channels due to their unique double-Bessel function structure. Concerning model performance, the Swin-Transformer model, benefiting from its efficient sliding-window self-attention mechanism, achieved a recognition accuracy of 98.2% under severe conditions of strong turbulence and 1 km transmission distance, significantly outperforming ResNet50 (94.8% ) and VGG16 (82% ). This study corroborates the practical application of BBG beams as highly robust information carriers. It validates the Swin-Transformer architecture’s superiority as an efficient demodulation tool under complex channel conditions, providing a novel solution for constructing high-capacity, highly reliable OAM-multiplexed FSO systems.
{"title":"Atmospheric Turbulence Transmission Characteristics and Deep Learning Identification of a Novel Bessel-Bessel-Gaussian Vortex Beam","authors":"Fu-Ming Ya;Xue-Jun Zhang;Pin-Pin Qin","doi":"10.1109/JPHOT.2026.3665803","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3665803","url":null,"abstract":"In the free-space optical communication (FSO) domain, orbital angular momentum (OAM) beams are vulnerable to disturbances from composite channels, including atmospheric turbulence and diffuse media, creating significant challenges for accurate mode identification. The present study investigates the transmission characteristics and mode recognition issues of a novel Bessel-Bessel-Gaussian (BBG) vortex beam after traversing a composite channel comprising atmospheric turbulence and diffuse media. Integrating the Fresnel diffraction model with the von Kalman turbulence model enables the numerical simulation of the optical field evolution of BBG beams under varying turbulence intensities and transmission distances. Simulation results in constructing a dataset comprising 9,600 distorted spot images. The evaluation of the recognition performance of multiple deep learning models, including traditional convolutional neural networks (VGG16, ResNet50) and advanced Transformer architectures (Swin-Transformer), is facilitated by this dataset. The simulation results demonstrate that BBG beams exhibit superior shape retention capabilities in complex channels due to their unique double-Bessel function structure. Concerning model performance, the Swin-Transformer model, benefiting from its efficient sliding-window self-attention mechanism, achieved a recognition accuracy of 98.2% under severe conditions of strong turbulence and 1 <italic>km</i> transmission distance, significantly outperforming ResNet50 (94.8% ) and VGG16 (82% ). This study corroborates the practical application of BBG beams as highly robust information carriers. It validates the Swin-Transformer architecture’s superiority as an efficient demodulation tool under complex channel conditions, providing a novel solution for constructing high-capacity, highly reliable OAM-multiplexed FSO systems.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-11"},"PeriodicalIF":2.4,"publicationDate":"2026-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11397676","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147362404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Raman spectroscopy as a non-destructive optical technique, is extensively employed for material characterization. However, inherently weak scattering signals and device-specific noises result in significant spectral variability across instruments, which severely limits the cross-device generalization of deep learning models. Since molecular vibrational signatures remain intrinsically device-invariant, we propose PhysRaman-Net, a novel physics-informed encoder-decoder framework for extracting device-invariant spectral representations. To mitigate the scarcity of labeled experimental data, we introduce a data augmentation strategy designed to model the noise differences among different Raman instruments. By modeling dominant noise sources according to their physical origins, we synthesize large-scale training samples that capture realistic spectral degradation patterns from theoretical spectra in the Computational Raman Database. Architecturally, we design a Permutation Invariant Encoder (PIE) based on the Transformer architecture to effectively capture the intrinsically unordered relationships among molecular vibrational modes. Distinctively, the proposed PhysDecoder is formulated as a differentiable physical model that explicitly encodes Raman scattering laws, thereby constraining the network to learn Raman tensors that represent the intensity information of Raman spectral peaks rather than abstract latent features. Extensive experiments on cross-device benchmarks demonstrate that PhysRaman-Net achieves superior generalization capability, attaining a classification accuracy of 96.8% for cross-device material classification, substantially outperforming state-of-the-art Convolutional Neural Network(CNN) and autoencoder baselines.
{"title":"PhysRaman-Net: A Physics-Informed Neural Network for Cross-Device Raman Spectral Analysis","authors":"Yilong Zhang;Jiawei Sun;Dongdong Zhao;Tianhao Lu;Ronghua Liang;Haohao Sun;Bangming Liu;Haixia Wang;Hui Zhang;Peng Chen","doi":"10.1109/JPHOT.2026.3662732","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3662732","url":null,"abstract":"Raman spectroscopy as a non-destructive optical technique, is extensively employed for material characterization. However, inherently weak scattering signals and device-specific noises result in significant spectral variability across instruments, which severely limits the cross-device generalization of deep learning models. Since molecular vibrational signatures remain intrinsically device-invariant, we propose PhysRaman-Net, a novel physics-informed encoder-decoder framework for extracting device-invariant spectral representations. To mitigate the scarcity of labeled experimental data, we introduce a data augmentation strategy designed to model the noise differences among different Raman instruments. By modeling dominant noise sources according to their physical origins, we synthesize large-scale training samples that capture realistic spectral degradation patterns from theoretical spectra in the Computational Raman Database. Architecturally, we design a Permutation Invariant Encoder (PIE) based on the Transformer architecture to effectively capture the intrinsically unordered relationships among molecular vibrational modes. Distinctively, the proposed PhysDecoder is formulated as a differentiable physical model that explicitly encodes Raman scattering laws, thereby constraining the network to learn Raman tensors that represent the intensity information of Raman spectral peaks rather than abstract latent features. Extensive experiments on cross-device benchmarks demonstrate that PhysRaman-Net achieves superior generalization capability, attaining a classification accuracy of 96.8% for cross-device material classification, substantially outperforming state-of-the-art Convolutional Neural Network(CNN) and autoencoder baselines.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-10"},"PeriodicalIF":2.4,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11386873","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In addition to information about cell morphology, hyperspectral images (HSIs) provide consistent spectral details across hundreds of wavelengths. Cytological analysis of samples extracted from serous cavity effusions is used to diagnose their benign or malignant origin but faces several challenges due to interference of various cell types. Therefore, cells segmentation from microscopy images is necessary to enable the extraction of specific quantitative biological characteristics. Here, we propose an automated semantic segmentation of cells into nucleus and cytoplasm classes using machine learning classifiers applied to HSIs. Moreover, the proposed procedure can be applied regardless of whether the cells are stained or unstained. We defined as input data multiple hyperspectral vectors derived from spectral profiles collected at the single-pixel scale from HSIs. The ensemble models achieving highest overall accuracy on the test sets were selected as demonstrators and subsequently applied under real conditions to the regions of interest of new, independent HSIs. The limitations associated with poor contrast in unstained samples were mitigated by introducing spectral shape transformations, along with a series of predefined post-processing functions, in order to improve the robustness of external evaluation. We achieved an overall accuracy of 0.98 for stained samples and 0.97 for unstained samples on the test sets. In the external evaluation, the Dice similarity coefficient was 0.94 for stained samples and 0.70 for unstained samples for the nucleus class, respectively, 0.88 and 0.82, for the cytoplasm class. This study advances the automatic segmentation methods for HSIs of stained and unstained samples. Future work will focus on integrating these findings into automated classification frameworks for cells harvested from serous cavity fluids.
{"title":"Hyperspectral Vectors for Machine Learning Segmentation of Stained and Unstained Cells From Serous Cavity Effusions","authors":"Mihaela-Andreea Ilisanu;Ana-Maria Pleava;Violeta Liuba Calin;Alina Ghioca;Elena Tianu;Valentin Popescu;Mihaela Georgeta Moisescu;Eugen Scarlat;Mona Mihailescu","doi":"10.1109/JPHOT.2026.3662426","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3662426","url":null,"abstract":"In addition to information about cell morphology, hyperspectral images (HSIs) provide consistent spectral details across hundreds of wavelengths. Cytological analysis of samples extracted from serous cavity effusions is used to diagnose their benign or malignant origin but faces several challenges due to interference of various cell types. Therefore, cells segmentation from microscopy images is necessary to enable the extraction of specific quantitative biological characteristics. Here, we propose an automated semantic segmentation of cells into nucleus and cytoplasm classes using machine learning classifiers applied to HSIs. Moreover, the proposed procedure can be applied regardless of whether the cells are stained or unstained. We defined as input data multiple hyperspectral vectors derived from spectral profiles collected at the single-pixel scale from HSIs. The ensemble models achieving highest overall accuracy on the test sets were selected as demonstrators and subsequently applied under real conditions to the regions of interest of new, independent HSIs. The limitations associated with poor contrast in unstained samples were mitigated by introducing spectral shape transformations, along with a series of predefined post-processing functions, in order to improve the robustness of external evaluation. We achieved an overall accuracy of 0.98 for stained samples and 0.97 for unstained samples on the test sets. In the external evaluation, the Dice similarity coefficient was 0.94 for stained samples and 0.70 for unstained samples for the nucleus class, respectively, 0.88 and 0.82, for the cytoplasm class. This study advances the automatic segmentation methods for HSIs of stained and unstained samples. Future work will focus on integrating these findings into automated classification frameworks for cells harvested from serous cavity fluids.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-10"},"PeriodicalIF":2.4,"publicationDate":"2026-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11373645","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-06DOI: 10.1109/JPHOT.2026.3662510
Victor Dmitriev;Cristiano Oliveira;Dimitrios C. Zografopoulos;Amanda Evangelista
We propose and theoretically investigate a new 2D structure consisting of monolayer phosphorene disk resonators in the unit cell of a square lattice. The crystallographic axes of anisotropic conductivity of the disks can be oriented differently with respect to the planes of symmetry of the square unit cell, which leads to different scattering matrices of the array. The first step is symmetry analysis of the array by group-theoretical methods and microwave circuit theory. Using the symmetry adapted linear combination method and the group approach, we discuss the principal characteristics of the eigenmodes classified by irreducible representations of the symmetry group of the resonator. Then, the numerical analysis of the array is undertaken in the regime of external excitation by incident planewaves. The calculations are carried out in the infrared frequency range. The principal new results of our work are: i) the calculated scattering matrices and their analysis, ii) tables of the eigenmodes with the currents and fields in the unit cell, iii) spectral analysis of the array for different excitation conditions, iv) investigated properties of the metasurface with protecting and supporting elements. The proposed structure can be used in infrared dynamically tunable filters, switches, modulators and sensors.
{"title":"Infrared Array With Circular Phosphorene Resonators in Square Unit Cells","authors":"Victor Dmitriev;Cristiano Oliveira;Dimitrios C. Zografopoulos;Amanda Evangelista","doi":"10.1109/JPHOT.2026.3662510","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3662510","url":null,"abstract":"We propose and theoretically investigate a new 2D structure consisting of monolayer phosphorene disk resonators in the unit cell of a square lattice. The crystallographic axes of anisotropic conductivity of the disks can be oriented differently with respect to the planes of symmetry of the square unit cell, which leads to different scattering matrices of the array. The first step is symmetry analysis of the array by group-theoretical methods and microwave circuit theory. Using the symmetry adapted linear combination method and the group approach, we discuss the principal characteristics of the eigenmodes classified by irreducible representations of the symmetry group of the resonator. Then, the numerical analysis of the array is undertaken in the regime of external excitation by incident planewaves. The calculations are carried out in the infrared frequency range. The principal new results of our work are: <italic>i)</i> the calculated scattering matrices and their analysis, <italic>ii)</i> tables of the eigenmodes with the currents and fields in the unit cell, <italic>iii)</i> spectral analysis of the array for different excitation conditions, <italic>iv)</i> investigated properties of the metasurface with protecting and supporting elements. The proposed structure can be used in infrared dynamically tunable filters, switches, modulators and sensors.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-8"},"PeriodicalIF":2.4,"publicationDate":"2026-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11373717","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223668","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Grating couplers that combine true vertical access, compact footprints, and foundry compatibility are vital for scalable photonic integration. However, conventional designs typically rely on reflectors, long tapers, or fiber tilt to achieve efficient coupling, which limits their practicality. Here, we present three inverse-designed vertical grating couplers on a standard 220 nm silicon-on-insulator platform without any bottom reflector in a single etch step, optimized through an adjoint-based topology framework with fabrication-aware filtering and projection enforcing a 90 nm feature constraint. The devices are tailored to three distinct fiber platforms: a single-polarization single-mode fiber, a dual-polarization single-mode fiber, and a hollow-core anti-resonant fiber. The resulting designs achieve efficient vertical coupling without the need for fiber tilt or tapering, while maintaining low crosstalk and robust performance across temperature variations and realistic alignment tolerances. These results establish inverse design as a viable and scalable approach for compact, broadband, and fabrication-ready vertical fiber-to-chip interfaces across diverse fiber types.
{"title":"Ultra-Compact Vertical Fiber-to-Chip Grating Couplers on Silicon-on-Insulator by Inverse Design","authors":"Jiahao Li;Xiang Li;Lin Wu;Hongguang Sun;Bohao Sun;Tiancai Jiang;Ming Luo;Zhan Gao;Yunlong Bai;Pengfei Ma;Yinqiu Gui;Jin Tao;Zhaofu Zhang;Hanbing Li;Tianye Huang;Ying Qiu","doi":"10.1109/JPHOT.2026.3661517","DOIUrl":"https://doi.org/10.1109/JPHOT.2026.3661517","url":null,"abstract":"Grating couplers that combine true vertical access, compact footprints, and foundry compatibility are vital for scalable photonic integration. However, conventional designs typically rely on reflectors, long tapers, or fiber tilt to achieve efficient coupling, which limits their practicality. Here, we present three inverse-designed vertical grating couplers on a standard 220 nm silicon-on-insulator platform without any bottom reflector in a single etch step, optimized through an adjoint-based topology framework with fabrication-aware filtering and projection enforcing a 90 nm feature constraint. The devices are tailored to three distinct fiber platforms: a single-polarization single-mode fiber, a dual-polarization single-mode fiber, and a hollow-core anti-resonant fiber. The resulting designs achieve efficient vertical coupling without the need for fiber tilt or tapering, while maintaining low crosstalk and robust performance across temperature variations and realistic alignment tolerances. These results establish inverse design as a viable and scalable approach for compact, broadband, and fabrication-ready vertical fiber-to-chip interfaces across diverse fiber types.","PeriodicalId":13204,"journal":{"name":"IEEE Photonics Journal","volume":"18 2","pages":"1-9"},"PeriodicalIF":2.4,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11372164","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：