You only train twice: A lighter and faster method for industrial weld defect detection based on dynamic kernel network
Xiaoyan Li, Liangliang Li, Peng Wang, Zhigang Lü, Hui Gao, Ruohai Di, Chuchao He, Chuchao He
Topics & Concepts
Computer scienceRandomnessData miningKernel (algebra)Scale (ratio)Field (mathematics)Artificial intelligenceContext (archaeology)AlgorithmMachine learningMathematicsBiologyPure mathematicsPhysicsCombinatoricsPaleontologyStatisticsQuantum mechanicsWelding Techniques and Residual StressesIndustrial Vision Systems and Defect DetectionNon-Destructive Testing Techniques