A5100625425- Orbital Angular Momentum in Optics
- Optical Systems and Laser Technology
- Advanced Fiber Laser Technologies
- Metamaterials and Metasurfaces Applications
- Neural Networks and Reservoir Computing
- Robot Manipulation and Learning
- Optical Network Technologies
- Adaptive optics and wavefront sensing
- Image Processing Techniques and Applications
- Digital Holography and Microscopy
- Advanced Antenna and Metasurface Technologies
- Microwave Engineering and Waveguides
- Advanced Image and Video Retrieval Techniques
- Multimodal Machine Learning Applications
- Optical measurement and interference techniques
- Advanced Memory and Neural Computing
- Electromagnetic Compatibility and Noise Suppression
- Gyrotron and Vacuum Electronics Research
- Flow Measurement and Analysis
- Domain Adaptation and Few-Shot Learning
- Advanced Sensor and Control Systems
- Antenna Design and Analysis
- Medical Imaging and Analysis
- Optical Wireless Communication Technologies
- Nonlinear Photonic Systems
Tsinghua University
2021-2024
Clemson University
2024
Qingdao University
2024
Beijing Institute of Technology
2013-2023
University of Shanghai for Science and Technology
2022
Beijing Normal University
2021
Peking University
2020
Peking University Third Hospital
2020
Structured light recently highlighted the higher-dimensional control with many tailored degrees of freedom (DoFs) such as amplitude, wavelength, polarization, and orbital angular momentum (OAM), spurring extension fundamental science advanced applications, especially novel information carrier to meet requirement everlastingly increasing communication bandwidth security. In addition basic DoFs in light, there are also several complex hidden spin–orbital coupled vector beams, ray-wave...
Abstract The structure and mechanism of the human visual system contain rich treasures, surprising effects can be achieved by simulating system. In this article, starting from system, we compare discuss discrepancies between traditional machine vision systems. Given wide variety large volume information, use non-von Neumann structured, flexible neuromorphic sensors effectively compensate for limitations systems based on von architecture. Firstly, article addresses emulation retinal...
Abstract Structured light with customized topological patterns inspires diverse classical and quantum investigations underpinned by accurate detection techniques. However, the current schemes are limited to vortex beams a simple phase singularity. The precise recognition of general structured multiple singularities remains elusive. Here, we report deep learning (DL) framework that can unveil multi-singularity structures in an end-to-end manner, after feeding only two intensity upon beam...
Optical vortex arrays (OVAs) have successfully aroused substantial interest from researchers for their promising prospects ranging classical to quantum physics. Previous reported OVAs still show a lack of controllable dimensions which may hamper applications. Taking an isolated perfect optical (POV) as array element, whose diameter is independent its topological charge (TC), this paper proposes combined phase-only holograms produce sophisticated POV arrays. The contributed scheme enables...
In light of pending capacity crunch in information era, orbital-angular-momenta-carrying vortex beams are gaining traction thanks to enlarged transmission capability. However, high-order confronted with fundamental limits nontrivial divergence or distortion, which consequently intensifies research on new optical states like low-order fractional beams. Here, we experimentally demonstrate an alternative mean increase the by simultaneously utilizing multiple non-orthogonal structured light,...
Photonic computing, with potentials of high parallelism, low latency and energy efficiency, have gained progressive interest at the forefront neural network (NN) accelerators. However, most existing photonic computing accelerators concentrate on discriminative NNs. Large-scale generative machines remain largely unexplored, partly due to poor data accessibility, accuracy hardware feasibility. Here, we harness random light scattering in disordered media as a native noise source leverage...
Optical microscopes have long been essential for many scientific disciplines. However, the resolution and contrast of such microscopic images are dramatically affected by aberrations. In this study, compacted with adaptive optics, we propose a machine learning technique, called 'phase-retrieval deep convolutional neural networks (PRDCNNs)'. This aberration determination architecture is direct exhibits high accuracy certain generalisation ability. Notably, its performance surpasses those...
Abstract For deep learning networks used to segment organs at risk (OARs) in head and neck (H&N) cancers, the class-imbalance problem between small volume OARs whole computed tomography (CT) images results delineation with serious false-positives on irrelevant slices unnecessary time-consuming calculations. To alleviate this problem, a slice classification model-facilitated 3D encoder–decoder network was developed validated. In two-step segmentation model, model firstly utilized classify...
Yb:CaGdAlO4, or Yb:CALGO, a new laser crystal, has been attracting increasing attention recently in myriad of technologies. This crystal features salient thermal, spectroscopic and mechanical properties, which enable highly efficient safe generation continuous-wave radiations ultrafast pulses with ever short durations. More specifically, its remarkable thermal-optic property high conversion efficiency allow high-power operation. Its nonlinear coefficient facilitates study optimized mode...
Softmax, a pervasive nonlinear operation, plays pivotal role in numerous statistics and deep learning (DL) models such as ChatGPT. To compute it is expensive especially for at-scale models. Several software hardware speed-up strategies are proposed but still suffer from low efficiency, poor scalability. Here we propose photonic-computing solution including massive programmable neurons that capable to execute operation an accurate, computation-efficient, robust scalable manner. Experimental...
Nonlinear optics of structured light carrying orbital angular momentum (OAM) has recently led to intriguing fundamental physical effects in light–matter interactions and advanced applications from classical imaging quantum informatics. However, the mutual interaction between spatial mode frequency conversions nonlinear processes is still elusive. In this work, we go beyond only considering OAM conversion investigate complex evolution cascaded astigmatic conversion. particular, unveil a...
In this paper, a WR-2.8 band (220-325 GHz) waveguide bandpass filter is presented, and it composed of five resonators. This improved directly coupled resonator with asymmetrical inductive windows realized using H-plane offsets. The Chebyshev response has centre frequency 300 GHz, fractional bandwidth 4.6% passband return loss 20 dB. Through the tolerance analysis, micro-electromechanical systems(MEMS) ultra-precision machining process precision within 5μm can be chosen for fabricating.
Untrained Physics-based Deep Learning (DL) methods for digital holography have gained significant attention due to their benefits, such as not requiring an annotated training dataset, and providing interpretability since utilizing the governing laws of hologram formation. However, they are sensitive hard-to-obtain precise object distance from imaging plane, posing Autofocusing challenge. Conventional solutions involve reconstructing image stacks different potential distances applying focus...
Human-robot interactions (HRIs) have attracted more and attentions during the past decades for fast conveniently introducing robot into human daily life. In this paper, one new human-robot cooperation method is developed, which falls in between full control autonomy. The operator charge of main operation autonomy gradually added to support execution operator's intent. proposed allows accomplish tasks effectively with help human. effectiveness verified by experiments, are based on Microsoft...
There is great necessity to allocate/combine the power of signal in present 340GHz microwave system. Therefore, this paper designed a directional coupler with branch waveguide structure. A five-branch structure was considering relative bandwidth and size After simulating analyzing model coupling by finite element method, we got result that it can realize 3dB during 325GHz~352GHz, while flatness less than 0.16dB. The phase difference two output ports 120 degrees.
Convolutional neural network has been widely used in hyperspectral image classification. Compared with the early machine learning method, it made great progress. However, convolution kernel classification ignores intrinsic relationship among spatial pixels when extracting spectral features, which will lead to poor contour and very small false prediction results. Besides, The data can only be labeled by experts, requires a lot of labor material resources. In order improve accuracy images...
High-precision detection of piston error is one the key technologies for high-resolution large-aperture segmented telescopes. Most methods based on neural networks are difficult to achieve high accuracy. In this Letter, we propose a high-precision method convolutional (CNN). A system with six sub-mirrors used, and set as reference mirror. The network can simultaneously extract information remaining five be tested from point spread function (PSF). training phase, 10,000 groups random values...