# GLipSDP
This repository contains the code used for the paper "Lipschitz constant estimation for general neural network architectures using control tools". The code is written in Matlab and uses Yalmip (https://yalmip.github.io/) and the solver Mosek (https://www.mosek.com/).
一般神经网络架构进行 Lipschitz 常数估计matlab代码.zip
版权申诉
91 浏览量
2024-05-18
14:29:55
上传
评论
收藏 91.41MB ZIP 举报
一般神经网络架构进行 Lipschitz 常数估计matlab代码.zip （70个子文件） 
一般神经网络架构进行 Lipschitz 常数估计matlab代码 GLipSDP-main Equivalence Tests 
test_pool.m 416B example_image.mat 3KB check_equivalence_conv_2.m 493B check_equivalence_conv.m 473B main_all.m 211B main_6C2F.m 2KB main_18ResNet_conv.m 2KB main_18ResNet_fc.m 1KB utils init_NN.m 148B ReLU.m 82B pool2fc.m 545B getRoesser.m 2KB LipschitzEstimationFazlyab.asv 690B conv2D_multi.m 520B LeNet2FC.m~ 604B conv2fc.m 1024B flatten.m 75B convert_weights.m 148B LipschitzEstimationFazlyab.m 710B simulate2DSysRoesser.m 775B LipEst.m 10KB main_2C2F.m 1KB main_4C3F.m 2KB main_sunetwork_tradeoff.m 1KB main_LeNet5_CIFAR.m 2KB Plots plot_subnetworks.m 253B plot_scalability.m 3KB scalability_subnetworks.tex 920B models weights_res_con.mat 177KB weights_cifar_6C2F_con.mat 9.89MB weights_mnist_LeNet5_con.mat 278KB weights_cifar_LeNet5_con.mat 278KB weights_mnist_4C3F_con.mat 8.72MB weights_l16_c16.mat 271KB weights_l16_c8.mat 101KB weights_mnist_2C2F_con.mat 746KB weights_res_fc_con.mat 523KB results mnist_ResNet_fc.mat 1.28MB mnist_main_6C2F.mat 10.01MB mnist_LeNet5.mat 2.84MB mnist_main_4C3F.mat 9.05MB mnist_ResNet_conv.mat 1.16MB subnetwork_tradeoff_8.mat 343KB mnist_main_2C2F.mat 4.89MB subnetwork_tradeoff.mat 1.21MB cifar_LeNet5.mat 4.4MB main_LeNet5_MNIST.m 2KB README.md 278B Weight conversion convert_6C2F.m 380B weights_mnist_2C2F.mat 651KB weights_res_con.mat 177KB convert_4C3F.m 337B weights_cifar_6C2F_con.mat 9.89MB weights_mnist_LeNet5_con.mat 278KB convert_LeNet5_cifar.m 257B weights_cifar_LeNet5_con.mat 278KB convert_ResNet_fc.m 1019B weights_mnist_4C3F_con.mat 8.72MB weights_mnist_4C3F.mat 7.53MB weights_cifar_6C2F.mat 8.59MB weights_l16_c16.mat 271KB weights_res_fc.mat 462KB convert_ResNet.m 978B convert_2C2F.m 216B convert_LeNet5.m 263B weights_mnist_LeNet5.mat 242KB weights_mnist_2C2F_con.mat 746KB weights_cifar_LeNet5.mat 243KB weights_res_fc_con.mat 523KB weights_res.mat 155KB资源评论
