Article
Chemistry, Analytical
Xiaoli Wang, Yan Piao, Jinyang Yu, Jie Li, Haixin Sun, Yuanshang Jin, Limin Liu, Tingfa Xu
Summary: This paper presents a novel Fourier ptychographic microscopy imaging reconstruction method based on a deep multi-feature transfer network, which achieves anti-noise performance, high-resolution, and reduced image data. By extracting image features using transfer learning ResNet50, Xception, and DenseNet121 networks, and adopting cascaded feature fusion strategy, as well as using pre-upsampling reconstruction network, high-quality image reconstruction is achieved.
Article
Optics
Famin Wang, Yun Xiao, Jiawang Zhao, Yunhai Zhang, Hangfeng Li
Summary: In this paper, a new method combining VSM, LSM, and FP algorithm is proposed for rapid super-resolution imaging. The iterative updating of low-resolution measurements taken under virtual modulation patterns leads to high-resolution and SNR image reconstruction.
OPTICS AND LASERS IN ENGINEERING
(2021)
Article
Biochemical Research Methods
Leena Thomas, M. K. Sheeja
Summary: Automated and accurate classification of breast cancer histological images is crucial for medical applications. This study combines Fourier ptychographic and deep learning to classify breast cancer histopathological images. The proposed technique generates high-resolution holograms and uses iterative retrieval with FP constraints to stitch low-resolution multi-view images. Feature extraction includes entropy, geometrical features, and textural features, with entropy-based normalization to optimize the features. The proposed ENDNN successfully classifies breast cancer images into normal or abnormal, outperforming traditional techniques.
JOURNAL OF BIOPHOTONICS
(2023)
Article
Engineering, Mechanical
Ruihua Liang, Weifeng Liu, Lihui Xu, Xiangyu Qu, Sakdirat Kaewunruen
Summary: This paper proposes a new PINN frequency domain (PINNFD) method by embedding Fourier features to solve the difficulty of approximating multi-frequency target functions in PINN. The effectiveness of the proposed method is validated by solving elastodynamics problems under various dynamic point loads. The results show that the proposed PINNFD method achieves better results in all cases of loading conditions, demonstrating its advancement in solving multi-frequency problems in engineering applications.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2023)
Article
Optics
An Pan, Aiye Wang, Junfu Zheng, Yuting Gao, Caiwen Ma, Baoli Yao
Summary: Fourier ptychographic microscopy (FPM) is a computational imaging technique that offers high resolution, wide field-of-view, and quantitative phase recovery. However, the presence of an imperceptible artifact caused by edge effect degrades the precision of phase imaging in FPM. To address this issue, two opposite algorithms called discrete cosine transform (DCT) and periodic plus smooth image decomposition (PPSID) were proposed and discussed systematically. The PPSID-FPM algorithm significantly improves the accuracy of phase measurement and is comparable to the conventional FPM algorithm.
OPTICS AND LASERS IN ENGINEERING
(2023)
Article
Quantum Science & Technology
Caio B. D. Goes, Thiago O. O. Maciel, Giovani G. G. Pollachini, Juan P. L. C. Salazar, Rafael G. G. Cuenca, Eduardo I. I. Duzzioni
Summary: This paper proposes a hybrid algorithm based on machine learning and quantum ensemble learning to find an approximate solution to a partial differential equation with good precision and favorable scaling in the required number of qubits. The classical component trains multiple regressors capable of approximately solving the equation using machine learning, while the quantum component adapts the QBoost algorithm to build an ensemble of classical learners. The algorithm is successfully applied to solve the 1D Burgers' equation with viscosity, demonstrating the improved solutions compared to classical weak-learners, and implemented on D-Wave Systems with good performance compared to other methods.
QUANTUM INFORMATION PROCESSING
(2023)
Article
Optics
Jinlei Zhang, Xiao Tao, Lin Yang, Chang Wang, Chenning Tao, Jing Hu, Rengmao Wu, Zhenrong Zheng
Summary: In this paper, a new Fourier Ptychographic Microscopy (FPM) method based on an untrained neural network (FuNN) is proposed to solve the FPM reconstruction problem by optimizing weights and bias through the interaction with a physical reconstruction model. Compared to other iterative FPM neural networks, FuNN can improve the quality of reconstructed complex object field without requiring a large number of training datasets.
OPTICS COMMUNICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Jianwen Gan, Longqing Zhang, Hongming Chen, Liping Bai, Xinwei Zhang, Lei Yang, Yanghong Zhang
Summary: The purpose of this study is to solve the problem of underground vehicle patrol by designing an unmanned patrol service using artificial intelligence technology. The design transforms a remote control car using Raspberry Pi development board, L298N driver chip, Raspberry Pi camera, and other major hardware equipment. Python is used as the programming language, with Python code controlling the car's movement through computer keyboard and collecting data with the camera. The Keras neural network library is utilized to quickly build and train a neural network model with the collected data. The model is then processed in the TensorFlow system to enable unmanned driving on a preset track.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Jizhou Zhang, Tingfa Xu, Yizhou Zhang, Yiwen Chen, Shushan Wang, Xin Wang
Summary: Fourier Ptychographic Microscopy (FPM) is a new technique for capturing wide field-of-view and high-resolution images suitable for the Internet of Medical Things (IoMT). The Multi-NNP framework simplifies the reconstruction process, improves performance, and promotes the application of high-resolution microscopic images in IoMT.
Article
Engineering, Chemical
Liqiang Lu, Xi Gao, Jean-Francois Dietiker, Mehrdad Shahnam, William A. Rogers
Summary: The research accelerated the computation speed of DEM using a convolutional neural network, demonstrating its accuracy and efficiency in the simulation of granular flows through experiments.
CHEMICAL ENGINEERING SCIENCE
(2021)
Article
Automation & Control Systems
Wen Song, Zhiguang Cao, Jie Zhang, Chi Xu, Andrew Lim
Summary: This paper proposes a deep reinforcement learning based approach to automatically discover new variable ordering heuristics for a given class of CSP instances. Experimental results show that the learned policies outperform classical hand-crafted heuristics in small and medium-sized instances, and also effectively reduce the search tree size in larger and harder instances.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Cell Biology
Aiye Wang, Zhuoqun Zhang, Siqi Wang, An Pan, Caiwen Ma, Baoli Yao
Summary: This paper introduces a method called ADMM-FPM which utilizes the concept of alternating direction method of multipliers to solve the phase retrieval problem in Fourier ptychographic microscopy (FPM). Compared to existing algorithms, ADMM-FPM shows better stability and robustness under noisy conditions.
Article
Computer Science, Information Systems
Lin Zhao, Xuhui Zhou, Xin Lu, Haiping Tong, Hui Fang
Summary: Fourier ptychographic microscopy (FPM) is a computational imaging technique that achieves high-resolution imaging with a simple setup. The FP-transformer, a transformer-based neural network, is proposed to reconstruct high-resolution FPM complex amplitude images from low-resolution amplitude (LRA) images. The FP-transformer demonstrates good performance in both training and validation, providing a dependable and adaptable platform for FPM deep learning reconstruction. The code for this work is available at https://github.com/zhaolin6/FPTransfomer for reproducibility purposes.
Article
Computer Science, Interdisciplinary Applications
Govinda Anantha Padmanabha, Nicholas Zabaras
Summary: Inverse modeling for computing a high-dimensional spatially-varying property field from sparse and noisy observations is a challenging problem, addressed by developing a surrogate model to estimate the unknown input field. The study constructs two-and three-dimensional inverse surrogate models for inversion tasks in multiphase flow problems.
JOURNAL OF COMPUTATIONAL PHYSICS
(2021)
Article
Nanoscience & Nanotechnology
Seungri Song, Jeongsoo Kim, Sunwoong Hur, Jaewoo Song, Chulmin Joo
Summary: PS-FPM combines high-resolution imaging with single-input-state illumination to achieve birefringence imaging of transparent objects without mechanical rotation, with a resolution of 0.55μm.
Article
Optics
He Zhang, Shaowei Jiang, Jun Liao, Junjing Deng, Jian Liu, Yongbing Zhang, Guoan Zheng
Article
Optics
Shaowei Jiang, Jun Liao, Zichao Bian, Pengming Song, Garrett Soler, Kazunori Hoshino, Guoan Zheng
Article
Optics
He Zhang, Zichao Bian, Shaowei Jiang, Jian Liu, Pengming Song, Guoan Zheng
Article
Optics
Xiu Li, Huaiyu Qi, Shaowei Jiang, Pengming Song, Guoan Zheng, Yongbing Zhang
Article
Optics
Teng Zhang, Shaowei Jiang, Zixin Zhao, Krishna Dixit, Xiaofei Zhou, Wa Hou, Yongbing Zhang, Chenggang Yan
Article
Optics
Pengming Song, Shaowei Jiang, He Zhang, Zichao Bian, Chengfei Guo, Kazunori Hoshino, Guoan Zheng
Article
Physics, Applied
Zichao Bian, Shaowei Jiang, Pengming Song, He Zhang, Pouria Hoveida, Kazunori Hoshino, Guoan Zheng
JOURNAL OF PHYSICS D-APPLIED PHYSICS
(2020)
Article
Optics
Chengfei Guo, Zichao Bian, Shaowei Jiang, Michael Murphy, Jiakai Zhu, Ruihai Wang, Pengming Song, Xiaopeng Shao, Yongbing Zhang, Guoan Zheng
Article
Optics
Pengming Song, Ruihai Wang, Jiakai Zhu, Tianbo Wang, Zichao Bian, Zibang Zhang, Kazunori Hoshino, Michael Murphy, Shaowei Jiang, Chengfei Guo, Guoan Zheng
Article
Optics
Chengfei Guo, Shaowei Jiang, Pengming Song, Zichao Bian, Tianbo Wang, Pouria Hoveida, Xiaopeng Shao
Summary: Diffractive zone plate optics use a micro-structure pattern to alter the propagation direction of incoming light waves, finding important applications in extreme-wavelength imaging. The resolution limit is determined by the smallest width of the outermost zone, and efforts have been made to fabricate very small zone widths for improved resolution. In an experiment, a diffractometer setup bypassed the resolution limit of zone plate optics, achieving 8-fold resolution improvement using in-plane and out-of-plane rotations as well as Fourier ptychographic image processing.
OPTICS COMMUNICATIONS
(2021)
Article
Biochemical Research Methods
Jun Liao, Xu Chen, Ge Ding, Pei Dong, Hu Ye, Han Wang, Yongbing Zhang, Jianhua Yao
Summary: This paper proposes a deep learning-based single-shot autofocus microscopy method that can achieve real-time and accurate autofocus. It supports the AI-based pathological diagnosis in digital pathology and has potential applications in life sciences, material research, and industrial automatic detection.
BIOMEDICAL OPTICS EXPRESS
(2022)
Article
Nanoscience & Nanotechnology
Shaowei Jiang, Chengfei Guo, Pengming Song, Niyun Zhou, Zichao Bian, Jiakai Zhu, Ruihai Wang, Pei Dong, Zibang Zhang, Jun Liao, Jianhua Yao, Bin Feng, Michael Murphy, Guoan Zheng
Summary: The article introduces a resolution-enhanced parallel coded ptychography technique that achieves high imaging throughput and the highest numerical aperture. The technique involves moving samples on a specially designed surface for imaging and collecting lensless diffraction data. A new coherent diffraction imaging model is proposed by considering the spatial and angular responses of pixel readouts.
Article
Optics
Shaoting Qi, Zilin Deng, Pan Qi, Jun Liao, Zibang Zhang, Guoan Zheng, Jingang Zhong
Summary: We report an active autofocusing method for sharp image projection, which works with wide-field structured illumination and single-pixel detection. The focus position is determined by recovering the Fourier coefficients from the single-pixel measurements and searching for the coefficient with the maximum magnitude. Experimental validation is performed and its effectiveness is demonstrated.
Article
Engineering, Electrical & Electronic
Xu Chen, Bowen Li, Shaowei Jiang, Terrance Zhang, Xu Zhang, Peiwu Qin, Xi Yuan, Yongbing Zhang, Guoan Zheng, Xiangyang Ji
Summary: DNN-SIM is a structured illumination microscopy framework powered by deep neural networks, which learns the physical relationship between images with different lateral phase shifts, reducing acquisition time significantly while maintaining resolution, and capable of handling new samples.
IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING
(2021)
Article
Biochemical Research Methods
Shaowei Jiang, Jiakai Zhu, Pengming Song, Chengfei Guo, Zichao Bian, Ruihai Wang, Yikun Huang, Shiyao Wang, He Zhang, Guoan Zheng