Results 141 to 150 of about 36,198 (253)

Multi‐Model Fruit Quality Assessment With Coarse YOLOv11 Presence Filtering, Vision Transformers and the Segment Anything Model

open access: yesApplied AI Letters, Volume 8, Issue 1, February 2027.
This paper presents a multi‐model deep learning pipeline for coarse fruit‐presence filtering, ripeness classification and pixel‐level spoilage segmentation. YOLOv11n is used only as an image‐level presence filter, four classifiers predict one of nine ripeness classes across three fruit types and the Segment Anything Model (SAM ViT‐H) generates fine ...
Md. Fazlay Rabbi   +4 more
wiley   +1 more source

Leakage‐Controlled Evaluation of Pixel‐Level MRI–sCT Fusion Reveals a Quality–Performance Mismatch in Brain Tumour Classification

open access: yesVector: The Journal of Engineering, Volume 2027, Issue 1, January/December 2027.
This study presents a leakage‐free evaluation of pixel‐level MRI–synthetic CT fusion for brain tumour classification using EfficientNet‐B0, image‐quality metrics, statistical testing and Grad‐CAM explainability. The results show that although fusion methods improved several image‐quality measures, none outperformed MRI‐only classification in accuracy ...
Muazzam Ali   +3 more
wiley   +1 more source

Combining Voiceprint Feature Extraction With Deep Learning Based Classifiers for Early Chronic Obstructive Pulmonary Disease Screening

open access: yesVector: The Journal of Engineering, Volume 2027, Issue 1, January/December 2027.
Graphic Abstract (GA) presents an overview of the CNN‐based classification process, which consists of several convolutional layers, several pooling layers, and a classification layer. Multilayer CNNs are capable of automatic feature enhancement, denoising, feature extraction, and pattern recognition tasks, and they have been applied in automatic ...
Pi‐Yun Chen   +6 more
wiley   +1 more source

EthioDance‐ViT: A Video Vision Transformer for Classifying Ethiopian Traditional Dances

open access: yesVector: The Journal of Engineering, Volume 2027, Issue 1, January/December 2027.
We present EthioDance‐ViT, a TimeSformer‐based Video Vision Transformer for classifying short video clips into five Ethiopian traditional dance categories Amhara, Oromo, Woliyta, Gurageya and Tigriga using a curated 18,060‐clip dataset spanning 1100 source videos. A key methodological contribution is a content‐based perceptual‐hash grouping that yields
Boaz B. Tulu   +5 more
wiley   +1 more source

GEP‐YOLO: A Lightweight Real‐Time Detection Model for Unmanned Aerial Vehicle Road Cracks Integrating Multi‐Module Optimization

open access: yesApplied Research, Volume 5, Issue 6, December 2026.
The graphical abstract summarizes GEP‐YOLO for UAV pavement crack detection, covering UAV data acquisition, dataset construction, and YOLOv11n optimization with GSConv, EUCB, and PReLU. These improvements enhance small‐crack recognition, real‐time performance, and robustness.
Ying Li   +4 more
wiley   +1 more source

Home - About - Disclaimer - Privacy