论文信息 - AdaGrad Gradient Descent Method for AI Image Management

AdaGrad Gradient Descent Method for AI Image Management

In this study, we used AdaGrad gradient descent method in optimizer for image deep learning, and compare with Adam gradient descent methods. After processing over six thousand huge database of through silicon via images, AdaGrad has shown a fast convergence and less generalization errors than Adam. The results help Artificial Intelligence for making the management of image judgment more accurate and faster.

[1] Peisheng Liu,et al. Advances in the Fabrication Processes and Applications of Wafer Level Packaging , 2014 .

[2] Trevor Darrell,et al. Fully Convolutional Networks for Semantic Segmentation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3] Ali Farhadi,et al. You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4] Yixin Yin,et al. Real-time Detection of Steel Strip Surface Defects Based on Improved YOLO Detection Network , 2018 .

[5] Geoffrey E. Hinton,et al. ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[6] H. Robbins. A Stochastic Approximation Method , 1951 .

[7] Thomas Luckenbach,et al. Automated inspection and metrology for 2.5D and 3D/TSV process assurance , 2014, 2014 IEEE 64th Electronic Components and Technology Conference (ECTC).

[8] Yoram Singer,et al. Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..