New State Space Model Enhances Power Transformer Defect Detection Accuracy
A new transformer defect detection method utilizing state space modeling and multi-scale feature fusion achieves an 84.7% mAP50 score, outperforming ...
A new transformer defect detection method utilizing state space modeling and multi-scale feature fusion achieves an 84.7% mAP50 score, outperforming ...
An integrated metrology framework combining multi-line 3D laser profilometry with an attention-enhanced U-Net achieved 97.3 percent accuracy in identifying microscopic ...
Researchers have created stochastic geometry models to synthesize realistic micro-scale textures for sandblasted and milled metallic surfaces, enhancing training data ...
A systematic review of seventy manufacturing studies reveals that deep learning architectures reach up to ninety-seven percent accuracy in defect ...
An adaptive hypergraph neural network integrating reinforcement learning outperformed baseline models by achieving up to 98.81 percent accuracy in rolling ...
Researchers have developed a parametric method using Voronoi tessellation to generate diverse synthetic defect geometries for training automated quality control ...
Researchers have demonstrated an automated structural health monitoring framework combining drone photogrammetry, laser scanning, and a lightweight neural network to ...
A dual-pathway AI framework combines impact-echo and ultrasonic signal analysis with CT imaging to detect hidden grouting defects and quantify ...
Engineers have developed a lightweight multi-scale neural network that uses component structural context to improve aerial defect detection across power ...
Researchers report an improved YOLO model that detects small PCB surface defects with 90.63% sensitivity while reducing model size and ...
© 2026 Scienmag - Science Magazine
© 2026 Scienmag - Science Magazine