Wangyu Wu

CAA-CLIP: Decoupled Pair and Cultural-Term Adaptation for Chinese Ancient Architecture Image–Text Retrieval

Xiaoyu Zhang, Hao Wang, Wangyu Wu, Qingqing Hu

Image–text retrieval for Chinese ancient architecture requires fine-grained discrimination among visually similar sites and alignment with specialist terms that are sparse in general vision–language corpora. We propose CAA-CLIP, a parameter-efficient framework for exact image–annotation retrieval and annotation-derived cultural-term retrieval. CAA-CLIP adapts five …

mo, cn, gb (code pays fourni par la source)

PCA–GPR-Assisted Sequential Bayesian Inversion of Slope Mechanical Parameters from Multi-Stage Deep Horizontal Displacement Monitoring

Youyun Li, Xu Chen, Zaiyang Yu, Wangyu Wu

Reliable mechanical parameters are needed to predict deformation during staged slope excavation, yet deep inclinometer profiles are high-dimensional and repeated numerical inversion is costly. This study integrates principal component analysis, Gaussian process regression (GPR), and sequential Bayesian updating to infer six effective …

cn, gb (code pays fourni par la source)

SDA-SwinNet: Swin-UNet with Dense Skip and Shift-ASPP for Retinal Vessel Segmentation

Jinghua Xiao, Ming Zhao, Rui Yang, Zhengqiang Wang et autres

Retinal artery/vein segmentation is a prerequisite for many ophthalmic diagnostic tools. Yet, the task remains difficult: vessels form complex trees, vary widely in caliber, and often appear low-contrast at terminal branches. We propose SDA-SwinNet to handle these challenges. The network adopts Swin-UNet …

cn, gb (code pays fourni par la source)

YOLO-PINE: A Lightweight Pineapple Object Detection Method Based on an Improved YOLO11n

Changxi Liu, Deji ZHAO, Jinbo Gao, Hang Shi et autres

Pineapple detection in complex field environments faces significant challenges, including severe background interference, frequent occlusion by sword leaves, and difficulties in identifying dense fruits, which demand an optimal balance between detection accuracy and computational efficiency for practical deployment on resource-constrained edge devices. …

cn, gb (code pays fourni par la source)

Mitigating Spectral Imbalance and Detail Attenuation in RGB-Thermal Object Detection via Frequency-Guided Multimodal Fusion

Quan Du, Ming Zhao, Song Lu, Minnan Hu et autres

RGB-T object detection combines visible texture information with thermal saliency cues to improve detection under degraded illumination. Existing RGB-T fusion methods usually perform feature interaction in the spatial domain or treat spectral responses jointly, which may allow coarse background components to dominate …

cn, gb (code pays fourni par la source)

Organic photonic synaptic transistors for bionic vision: From spectral engineering to flexible intelligent perception

Longlong Jiang, Shan Wu, Silu Feng, H Huang et autres

Abstract Bionic intelligent visual perception systems have significant application prospects in fields such as autonomous driving and robotics, medical imaging and auxiliary diagnosis, security and monitoring. Achieving spectral perception through material design and device structure adjustment is the core foundation for building …

cn, sg, gb, us (code pays fourni par la source)

Federated Multi-View Unsupervised Feature Selection via Bio-Inspired Hierarchical-Cognitive Tianji’s Horse Racing Optimization and Tensor Learning

Rong Cheng, Zhiwei Sun, Kun Qi, Wangyu Wu et autres

As multi-view datasets expand across diverse practical fields, feature selection (FS) has become an indispensable preparatory stage for machine learning models. Nevertheless, real-world multi-view data is often unlabeled and distributed among isolated clients, posing significant challenges to traditional centralized methods due to …

cn, gb (code pays fourni par la source)

A Training-Free Paradigm for Data-Scarce Maritime Scene Classification Using Vision-Language Models

Jiabao Wu, Yujie Chen, Wentao Chen, Y. M. Lai et autres

Maritime Domain Awareness (MDA) relies heavily on data acquired from high-resolution optical spaceborne sensors; however, processing this massive quantity of sensor data via traditional supervised deep learning is severely bottlenecked by its dependency on exhaustively annotated datasets. Under extreme data scarcity, conventional …

cn, gb (code pays fourni par la source)

M2Caller: A Mamba2 Based Oxford Nanopore Basecaller Balancing Accuracy and Throughput

Jie Qiu, Wangyu Wu, H Li, Haslina Binti Hashim

Basecalling transforms the noisy electrical signals generated during nanopore sequencing into nucleotide sequences, and its accuracy critically determines the reliability of downstream analyses. Increasingly complex architectures have been adopted for basecalling, yet this trend has produced bloated systems that are difficult to …

my, gb, us (code pays fourni par la source)

Enhancing Small Cell Detection in Whole Slide Images with Saliency-Guided Transformers

Linkun Yu, Yuehua Chen, Qianxi Yan, Wendi Rao et autres

Detecting inflammatory cells in whole slide images (WSIs) is crucial for kidney transplant assessment but remains challenging due to their small size and dense distribution. We present the Structured Saliency-Guided Transformer (SSGT), an extension of Deformable DETR for small-object detection. SSGT integrates …

jp, cn, sg, gb (code pays fourni par la source)

LLM-enhanced multimodal fusion for cross-domain sequential recommendation

Wangyu Wu, Z H Chen, Wenqiao Zhang, Siqi Song et autres

Interpretable Sensor Change Detection via Conditional Cauchy–Schwarz Divergence

Wenyu Wang, Yuan Shen, Yao Ni, Wangyu Wu

Detecting distributional changes in multivariate sensor networks is a fundamental task for monitoring complex systems such as industrial processes, structural health monitoring, and large-scale Internet of Things infrastructures. Despite significant progress, most existing change-point detection methods either operate on high-dimensional observations in …

gb, us, cn (code pays fourni par la source)