Jiayu Yang

ORCID: 0000-0001-5102-1002
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Photonic and Optical Devices
  • Photorefractive and Nonlinear Optics
  • Video Coding and Compression Technologies
  • Advanced Data Compression Techniques
  • Advanced Fiber Laser Technologies
  • Advanced Image Processing Techniques
  • Advanced Vision and Imaging
  • Optical Network Technologies
  • Image Retrieval and Classification Techniques
  • Video Analysis and Summarization
  • Smart Agriculture and AI
  • Neural Networks and Reservoir Computing
  • Industrial Vision Systems and Defect Detection
  • Supercapacitor Materials and Fabrication
  • Energy, Environment, Economic Growth
  • Advanced battery technologies research
  • Quantum Information and Cryptography
  • Vehicular Ad Hoc Networks (VANETs)
  • Fermentation and Sensory Analysis
  • Security in Wireless Sensor Networks
  • Power Transformer Diagnostics and Insulation
  • Advanced Image and Video Retrieval Techniques
  • Power Systems Fault Detection
  • Transportation Systems and Logistics
  • E-commerce and Technology Innovations

Southwest Forestry University
2025

University of Science and Technology of China
2022-2025

Ningbo Polytechnic
2024

Northwestern Polytechnical University
2023-2024

Zhejiang University
2023-2024

Guilin University of Electronic Technology
2023-2024

Peking University
2020-2024

Xinjiang University
2024

National Taiwan University
2024

Beijing University of Posts and Telecommunications
2018-2024

By integrating the traditional power grid with information and communication technology, smart achieves dependable, efficient, flexible data processing. The meters deployed on user side of collect users' usage a regular basis upload it to control center complete acquisition. can evaluate supply demand through aggregated from users then dynamically adjust price, etc. However, since collected may disclose user's electricity habits daily activities, privacy concern has become critical issue in...

10.1016/j.dcan.2022.01.004 article EN cc-by-nc-nd Digital Communications and Networks 2022-01-26

The basic indexes of all-optical integrated photonic circuits include high-density integration, ultrafast response and ultra-low energy consumption. Traditional methods mainly adopt conventional micro/nano-structures. overall size the circuit is large, usually reaches hundreds microns. Besides, it difficult to balance consumption problem, crosstalk between two traditional devices overcome. Here, we propose experimentally demonstrate an approach based on inverse design method realize a...

10.29026/oea.2022.210061 article EN cc-by Opto-Electronic Advances 2022-01-01

<title>Abstract</title> Recent advancements in thin-film lithium niobate (TFLN) photonics have led to a new generation of high-performance electro-optic devices, including modulators, frequency combs, and microwave-to-optical transducers. However, the broader adoption TFLN-based devices that rely on all-optical nonlinearities been limited by sensitivity quasi-phase matching (QPM), realized via ferroelectric poling, fabrication tolerances. Here, we propose scalable process aimed at improving...

10.21203/rs.3.rs-4289238/v1 preprint EN cc-by Research Square (Research Square) 2024-05-07

Warehouses are critical logistics nodes, with food freezer warehouses playing a key role in ensuring quality while facing challenges such as high-density item distribution and extremely low temperatures required for occupational safety. Traditional management methods struggle to meet these demands, underscoring the need intelligent digital solutions improve efficiency mitigate safety risks. This study proposes YOLOv8-RSS model, lightweight high-precision approach tailored warehouse...

10.1038/s41598-025-86662-z article EN cc-by-nc-nd Scientific Reports 2025-01-17

All-optical switches are among the most important parts of integrated photonics. Ultrahigh speed and ultralow energy consumption two necessary indexes all-optical switches. Traditionally, based on concepts such as micro-ring resonators, surface plasmon polaritons, photonic crystals, metamaterials. However, platforms cannot satisfy demand for high performance To overcome limited response time consumption, recent studies have introduced new applications physics parity–time symmetry,...

10.1063/5.0048878 article EN Journal of Applied Physics 2021-06-04

Flexible solid-state Zn-ion batteries (ZIBs) have garnered considerable attention for next-generation power sources, but the corrosion, dendrite growth, and interfacial problems severely hinder their practical applications. Herein, a high-performance flexible ZIB with unique heterostructure electrolyte is facilely fabricated through ultraviolet-assisted printing strategy. The solid polymer/hydrogel matrix not only isolates water molecules optimizes electric field distribution dendrite-free...

10.1002/smll.202303108 article EN Small 2023-05-24

Recent advancements in thin-film lithium niobate (TFLN) photonics have led to a new generation of high-performance electro-optic devices, including modulators, frequency combs, and microwave-to-optical transducers. However, the broader adoption TFLN-based devices that rely on all-optical nonlinearities been limited by sensitivity quasi-phase matching (QPM), realized via ferroelectric poling, fabrication tolerances. Here, we propose scalable process aimed at improving wavelength-accuracy...

10.48550/arxiv.2404.12381 preprint EN arXiv (Cornell University) 2024-04-18

Ensuring a consistently reliable power supply is paramount in systems. Researchers are engaged the pursuit of categorizing transmission line failures to design countermeasures for mitigating associated financial losses. Our study employs machine learning-based methodology, specifically Conformer Convolution-Augmented Transformer model, classify fault types. This model processes time series input data directly, eliminating need expert feature extraction. The training and validation datasets...

10.3390/app14104031 article EN cc-by Applied Sciences 2024-05-09

Adversarial mechanism is introduced to learned image compression system in this paper. Our motivation that the number of quantization levels limited with constraint low bit-rate, resulting severe distortion details after reconstruction. The adversarial training manner enhances ability Decoder/Generator enrich textures and reconstructed image. Channel-spatial attention used refine intermediate features implicitly boost representation power CNNs. As for entropy model, we jointly take...

10.1109/cvprw50498.2020.00078 article EN 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) 2020-06-01

Abstract The development of information technology urgently requires ultrafast, ultra-low energy consumption and ultra-high-capacity data computing abilities. Traditional method electronic chips is limited by the bottleneck Moore’s Law. All-optical photonic provides a promising way to realize such high-performance Until now, it still huge challenge all-optical four arithmetic operations at same time on chip. Here, we propose new encoding scheme for binary computation, including n -bit...

10.1515/nanoph-2021-0467 article EN cc-by Nanophotonics 2021-10-26

A Quantum secure direct communication (QSDC) and authentication protocol based on the W-class state is presented to enhance efficiency of eavesdropping detection. In this protocol, used transmission checking sequence detect eavesdroppers. security analysis, method entropy theory introduced, two detection strategies are compared quantitatively by using constraint between information that eavesdroppers can obtain a normalized difference parameter introduced. To same amount information,...

10.1049/cje.2017.10.006 article EN Chinese Journal of Electronics 2018-03-01

Learned video compression has developed rapidly and shown promising rate-distortion performance recently. Existing works have made great progress on removing temporal redundancy between inter-frames, while neglecting spatial within a frame. In this paper, we propose to explore further improve efficiency reduce computation complexity by extending inter-frame prediction from spatial-temporal prediction. Specifically, the current frame is first down-sampled redundancy, which then temporally...

10.1109/icassp48485.2024.10448496 article EN ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2024-03-18

In neural video codecs, current state-of-the-art methods typically adopt multi-scale motion compensation to handle diverse motions. These estimate and compress either optical flow or deformable offsets reduce inter-frame redundancy. However, flow-based often suffer from inaccurate estimation in complicated scenes. Deformable convolution-based are more robust but have a higher bit cost for coding. this paper, we propose hybrid context generation module, which combines the advantages of above...

10.1109/dcc58796.2024.00040 article EN 2024-03-19

We report progress on a process flow for fabricating wavelength-accurate quasi-phasematched frequency conversion devices in thin-film lithium niobate. Of 156 sampled devices, ~48% are phasematched within ±5 nm of target second-harmonic generation process.

10.1364/cleo_si.2024.stu3e.2 article EN 2024-01-01

The ferroelectric domains of thin-film lithium niobate are structured for multi-wavelength nonlinear optical frequency conversion. We target frequency-multiplexed photon-pair generation, in which we show conditions five pair sources (purity &gt; 90%) a single waveguide.

10.1364/cleo_fs.2024.ff2k.7 article EN 2024-01-01

Currently, the high-speed performance of thin-film lithium niobate electro-optic modulator chips is evolving rapidly. Nevertheless, due to inherent technical limitations imposed by packaging design and material architecture, intrinsic bandwidth often exceeds their interfaces, which can constrain realization modulation performance. Bump bonding emerges as a high-bandwidth EO interconnection technology, outperforming wire for faster optical communication. In this paper, we present chip...

10.3390/electronics13224463 article EN Electronics 2024-11-14

In neural video codecs, current state-of-the-art methods typically adopt multi-scale motion compensation to handle diverse motions. These estimate and compress either optical flow or deformable offsets reduce inter-frame redundancy. However, flow-based often suffer from inaccurate estimation in complicated scenes. Deformable convolution-based are more robust but have a higher bit cost for coding. this paper, we propose hybrid context generation module, which combines the advantages of above...

10.48550/arxiv.2412.00446 preprint EN arXiv (Cornell University) 2024-11-30
Coming Soon ...