MYMengxiao Yin

Mengxiao Yin

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Mengxiao Yin has a total of 45 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Intelligent Design of a Wavelength-Sensitive Metalens for Image Encryption

    by Haoyu Cheng, Haitao Chen, Mengxiao Yin, Zhenrong Zhang

    As a cutting-edge subwavelength diffractive optical device, the metalenses demonstrate significant application potential in fields such as high-resolution imaging and optical communications, and their design methodologies are advancing into a new phase with the development of intelligent technologies like neural networks. In response to the growing demands for information security and the challenges faced by digital encryption in the context of high-performance computing and quantum attacks, optical encryption technology has emerged as a highly promising alternative, leveraging its ability to manipulate high-dimensional physical parameters and inherent parallelism. This paper proposes an image encryption method based on wavelength-sensitive metalenses, integrating deep learning and optimization algorithms to construct a highly secure optical encryption system capable of jointly modulating the dual-wavelength ratio and polarization state. In this study, a multilayer perceptron (MLP) surrogate model was constructed to replace computationally expensive finite-difference time-domain (FDTD) simulations, thereby enabling rapid and differentiable prediction of the optical responses of titanium dioxide (TiO2) nanopillar units. Subsequently, by combining the particle swarm optimization (PSO) algorithm, a highly dispersive metalens was designed, which achieves high-quality imaging only under specific conditions: incident light with a precise ratio of dual wavelengths (550 nm and 450 nm) and a 45°linear polarization state. The point spread function (PSF) of this lens serves as the optical key. During the verification phase, image encryption and decryption were accomplished using image convolution and Wiener deconvolution. Tests on 4000 random samples demonstrated that the system exhibits nearly complete decryption failure under illegal keys, highlighting its exceptional encryption performance and robustness. By utilizing the dual-wavelength ratio as the core physical key, this work provides novel design insights and practical pathways for the intelligent design of meta-optical devices and high-dimensional optical encryption.

  • SAGA-Feat: A semantic- and geometry-aware network for sparse local feature learning

    by Yanhan Mo, Mengxiao Yin, Guiqing Li, Junjie Liao, Zhijie Liang

  • Open Access

    FFTNet: Fusing Frequency and Temporal Awareness in Long-Term Time Series Forecasting

    by Zhiqiang Yang, Mengxiao Yin, Junjie Liao, Fancui Xie, Peizhao Zheng, Jiachao Li, Bei Hua

    Time series forecasting is extensively utilised in meteorology, transportation, finance, and industrial domains. Precisely recognising cyclical trends and abrupt local changes in time series is essential for enhancing forecasting accuracy. Frequency-domain representations are adept at identifying periodic traits, whereas time-domain approaches are superior for spotting localised quick changes. Traditional techniques sometimes prioritise a single domain, overlooking the advantages of integration. This paper introduces a novel hybrid model, FFTNet, that simultaneously pulls characteristics from both domains to optimise their respective benefits. Theoretical examination indicates that the 2D CNN utilises dual-axis convolution kernels to jointly describe global cross-patch structures and local temporal patterns, whilst the frequency-domain MLP enhances spectral components in accordance with Parseval’s Theorem and the Convolution Theorem. A frequency-domain MLP is employed to discern periodic and trend characteristics, while a 2D CNN in the time domain identifies localised abrupt changes. This hybrid methodology differentiates itself from previous methods that depend exclusively on a single domain, providing a more thorough comprehension of the underlying patterns in time series data. Experiments on seven real-world datasets indicate that FFTNet surpasses existing techniques, attaining state-of-the-art performance with enhancements of 11.8% and 4.7% in MSE and MAE, respectively.


  • ALFC-Point: Adaptive Laplacian Feature Convolution Network for 3D Point Cloud Understanding

    by Junjie Liao, Mengxiao Yin, Ming Li, Congyang Zhu, Zhiqiang Yang, Feng Zhan

  • Open Access

    Three-Dimensional Mesh Character Pose Transfer with Neural Sparse-Softmax Skinning Blending

    by Siqi Liu, Mengxiao Yin, Ming Li, Feng Zhan, Bei Hua

    Three-dimensional mesh pose transfer transforms the pose of a source model into the pose of a reference model while preserving the source model’s identity (body detail). It has tremendous potential in computer graphics tasks. Current neural network-based methods primarily focus on extracting pose and body features, not entirely using the articulated body structure of humans and animals. We propose an end-to-end pose transfer network based on skinning deformation to address these issues. This network first extracts skinning weights and model joint features. Then, they are decoded to transfer the source model to a pose similar to the reference model while preserving the features of the source model. During feature extraction, we utilize the features of the k-nearest neighborhoods and one-ring neighborhoods to enable the network to learn the body details of the model better. Additionally, we apply skinning weights and joint features to capture the variation in the source model pose compared to the reference model pose and then use a decoding network to obtain the target model, replacing linear blend skinning. We conducted experiments on datasets such as SMPL, SMAL, FAUST, DYNA, and the MG dataset to provide empirical evidence and demonstrate that our method is the best in quantitative experiments. Our method efficiently transfers poses while better preserving the identity of the source model.

  • Open Access

    DW-MLSR: Unsupervised Deformable Medical Image Registration Based on Dual-Window Attention and Multi-Latent Space

    by Yuxuan Huang, Mengxiao Yin, Zhipan Li, Feng Yang

    (1) Background: In recent years, the application of Transformers and Vision Transformers (ViTs) in medical image registration has been constrained by sliding attention mechanisms, which struggle to effectively capture non-adjacent but critical structures, such as the hippocampus and ventricles in the brain. Additionally, the lack of labels in unsupervised registration often leads to overfitting. (2) To address these issues, we propose a novel method, DW-MLSR, based on dual-window attention and multi-latent space. The dual-window attention mechanism enhances the transmission of information across non-adjacent structures, while the multi-latent space improves the model’s generalization by learning latent image representations. (3) Experimental results demonstrate that DW-MLSR outperforms mainstream registration models, showcasing significant potential in medical image registration. (4) The DW-MLSR method addresses the limitations of sliding attention in transmitting information between non-adjacent windows, improves the performance of unsupervised registration, and demonstrates broad application prospects in medical image registration.


  • Object and spatial discrimination makes weakly supervised local feature better

    by Yifan Yin, Mengxiao Yin, Yunhui Xiong, Pengfei Lai, Kan Chang, Feng Yang