Ashish Patel 🇮🇳’s Post

Day-27 Computer Vision Learning AmoebaNet — Regularized Evolution for Image Classifier Architecture Search (Image Classification) by Google Brain Follow me for similar post : 🇮🇳 Ashish Patel 𝗜𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝗶𝗻𝗴 𝗙𝗮𝗰𝘁𝘀 : 🔸 It is published in 2019 AAAI, which has already got over 1013 citations 🔸 Image Classifier AmoebaNet-A, is evolved, by evolutionary algorithm, that surpasses hand-designs for the first time. 🔸 The tournament selection evolutionary algorithm is modified by introducing an age property to favor the younger genotypes. 🔸 83.9% top-1 / 96.6% top-5 ImageNet accuracy is obtained. ------------------------------------------------------------------- 𝗔𝗺𝗮𝘇𝗶𝗻𝗴 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 : https://lnkd.in/eEZPMSh official tensorflow : https://bit.ly/3t3ssUC Keras and tensorflow : https://bit.ly/3t2mXWb Pytorch : https://bit.ly/3t0fqab Colab : https://bit.ly/3a9sFg6 ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 AmoebaNet is used Search Space, Aging Evolutionary Algorithm 🔸 The Goal of Architecture Search Process is to discover the architectures of normal and reduction cells. 🔸 Normal cell Extract the features and Reduction cell down-sample the features. #artificialintelligence #computervision #innovation