vgg19
paddle.vision.models.vgg19
( pretrained=False, batch_norm=False, \*kwargs* ) [源代码]
vgg19模型,来自论文 “Very Deep Convolutional Networks For Large-Scale Image Recognition” 。
参数
pretrained (bool,可选) - 是否加载在imagenet数据集上的预训练权重。默认值:False。
batch_norm (bool, 可选) - 是否在每个卷积层后添加批归一化层。默认值:False。
返回
vgg19模型,Layer的实例。
代码示例
import paddle
from paddle.vision.models import vgg19
# build model
model = vgg19()
# build vgg19 model with batch_norm
model = vgg19(batch_norm=True)
x = paddle.rand([1, 3, 224, 224])
out = model(x)
print(out.shape)