2026a
# DCGANdiscriminator
深层卷积生成对抗网络判别器
函数库: TyDeepLearning
# 语法
net = DCGANdiscriminator(imageSize, numChannel)
# 说明
net = DCGANdiscriminator(imageSize, numChannel)创建深层卷积生成对抗网络判别器。
# 示例
创建深层卷积生成对抗网络判别器
创建一个深层卷积生成对抗网络判别器。
using TyDeepLearning
set_backend(:mindspore)
net = DCGANdiscriminator(64, 3)
PyObject Discriminator<
(discriminator): SequentialCell<
(0): Conv2d<input_channels=3, output_channels=64, kernel_size=(4, 4), stride=(2, 2), pad_mode=pad, padding=1, dilation=(1, 1), group=1, has_bias=False, weight_init=<mindspore.common.initializer.Normal object at 0x000002383C41DF08>, bias_init=None, format=NCHW>
(1): LeakyReLU<>
(2): Conv2d<input_channels=64, output_channels=128, kernel_size=(4, 4), stride=(2, 2), pad_mode=pad, padding=1, dilation=(1, 1), group=1, has_bias=False, weight_init=<mindspore.common.initializer.Normal object at 0x000002383C41DF08>, bias_init=None, format=NCHW>
(3): BatchNorm2d<num_features=128, eps=1e-05, momentum=0.9, gamma=Parameter (name=discriminator.3.gamma, shape=(128,), dtype=Float32, requires_grad=True), beta=Parameter (name=discriminator.3.beta, shape=(128,), dtype=Float32, requires_grad=True), moving_mean=Parameter (name=discriminator.3.moving_mean, shape=(128,), dtype=Float32, requires_grad=False), moving_variance=Parameter (name=discriminator.3.moving_variance, shape=(128,), dtype=Float32, requires_grad=False)>
(4): LeakyReLU<>
(5): Conv2d<input_channels=128, output_channels=256, kernel_size=(4, 4), stride=(2, 2), pad_mode=pad, padding=1, dilation=(1, 1), group=1, has_bias=False, weight_init=<mindspore.common.initializer.Normal object at 0x000002383C41DF08>, bias_init=None, format=NCHW>
(6): BatchNorm2d<num_features=256, eps=1e-05, momentum=0.9, gamma=Parameter (name=discriminator.6.gamma, shape=(256,), dtype=Float32, requires_grad=True), beta=Parameter (name=discriminator.6.beta, shape=(256,), dtype=Float32, requires_grad=True), moving_mean=Parameter (name=discriminator.6.moving_mean, shape=(256,), dtype=Float32, requires_grad=False), moving_variance=Parameter (name=discriminator.6.moving_variance, shape=(256,), dtype=Float32, requires_grad=False)>
(7): LeakyReLU<>
(8): Conv2d<input_channels=256, output_channels=512, kernel_size=(4, 4), stride=(2, 2), pad_mode=pad, padding=1, dilation=(1, 1), group=1, has_bias=False, weight_init=<mindspore.common.initializer.Normal object at 0x000002383C41DF08>, bias_init=None, format=NCHW>
(9): BatchNorm2d<num_features=512, eps=1e-05, momentum=0.9, gamma=Parameter (name=discriminator.9.gamma, shape=(512,), dtype=Float32, requires_grad=True), beta=Parameter (name=discriminator.9.beta, shape=(512,), dtype=Float32, requires_grad=True), moving_mean=Parameter (name=discriminator.9.moving_mean, shape=(512,), dtype=Float32, requires_grad=False), moving_variance=Parameter (name=discriminator.9.moving_variance, shape=(512,), dtype=Float32, requires_grad=False)>
(10): LeakyReLU<>
(11): Conv2d<input_channels=512, output_channels=1, kernel_size=(4, 4), stride=(1, 1), pad_mode=valid, padding=0, dilation=(1, 1),
group=1, has_bias=False, weight_init=<mindspore.common.initializer.Normal object at 0x000002383C41DF08>, bias_init=None, format=NCHW>
>
(adv_layer): Sigmoid<>
>
# 输入参数
imageSize-生成图像尺寸标量
生成的正方形图像尺寸，代表图像的高和宽。
数据类型： Int64
numChannel-通道数向量
生成图像的通道数。
数据类型： Int64
# 输出参数
net-深层卷积生成对抗网络判别器网络对象
输出为一个深层卷积生成对抗网络判别器。