pytorch实现黑盒攻击

最新推荐文章于 2025-01-21 17:45:00 发布

原创

最新推荐文章于 2025-01-21 17:45:00 发布 · 365 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#pytorch #人工智能

源码链接：https://github.com/wanglouis49/pytorch-adversarial_box
在LeNet5上进行黑盒攻击
FGSM生成对抗样本

substituteModel上训练

def MNIST_bbox_sub(param, loader_hold_out, loader_test):

    # Setup training
    optimizer = torch.optim.Adam(net.parameters(), lr=param['learning_rate'])

    # Data held out for initial training
	# .....

    # 训练
    for rho in range(param['data_aug']):
        print("Substitute training epoch #"+str(rho))
        print("Training data: "+str(len(X_sub)))

        rng = np.random.RandomState