使用训练好的权重并改进网络结构提取特征

小菜白找码码

已于 2023-04-13 17:45:47 修改

阅读量181

点赞数

文章标签：深度学习 numpy python

于 2023-04-12 21:13:48 首次发布

本文链接：https://blog.csdn.net/baseclanguage/article/details/130116892

版权

    # create model
    model = create_model(num_classes=num_classes)
    model.build((32, img_size, img_size, 3))

    # 下载我提前转好的预训练权重
    # 链接: https://pan.baidu.com/s/1cHVwia2i3wD7-0Ueh2WmrQ  密码: sq8c
    # load weights
    pre_weights_path = './swin_small_patch4_window7_224.h5'
    assert os.path.exists(pre_weights_path), "cannot find {}".format(pre_weights_path)
    model.load_weights(pre_weights_path, by_name=True, skip_mismatch=True)

    # freeze bottom layers
    if freeze_layers:
        for layer in model.layers:
            if "head" not in layer.name:
                layer.trainable = False
            else:
                print("training {}".format(layer.name))

    model.summary()
    input = np.random.rand(32, 244, 244, 3)
    print(input.shape)
    out = model(input)
    print(out.shape)

net_size_image = cv2.resize(rgb_image, (img_size, img_size))
    net_image = tf.expand_dims(net_size_image, axis=0)
    net_image = tf.cast(net_image,dtype=tf.float32)
    tf_net_image = tf.convert_to_tensor(net_image, dtype=tf.float32)
    output = model(tf_net_image, training=False)
    output = np.array(output)
    output = output.flatten()
    print(output)