Python Keras.net预测SSD模型和返回数据阵列问题

Python Keras.net预测SSD模型和返回数据阵列问题,python,c#,numpy,keras,Python,C#,Numpy,Keras,我使用Keras.net(c#)并加载SSD模型() 尝试预测lenna.jpg图像。但当预测到时,我得到了返回值(NDarray),我尝试从NDarray获取一个对象,然后返回错误 NDArray值: NDarray result = m_Model.Predict(image_exp); [array([[[ 0.17123657, 0.18411542, 0.2141329 , 0.89467835], [ 0.81485975, 1.0284375 , -0.4

我使用Keras.net(c#)并加载SSD模型() 尝试预测lenna.jpg图像。但当预测到时,我得到了返回值(NDarray),我尝试从NDarray获取一个对象,然后返回错误

NDArray值:

NDarray result = m_Model.Predict(image_exp);

[array([[[ 0.17123657,  0.18411542,  0.2141329 ,  0.89467835],
        [ 0.81485975,  1.0284375 , -0.41349524,  0.6665468 ],
        [ 0.01628919,  0.11503349,  0.8445237 ,  0.06173617],
        ...,
        [-1.1167231 , -1.4015226 , -2.944187  , -0.9941753 ],
        [-1.4996854 , -1.6622332 , -2.0316055 , -1.8165188 ],
        [-1.4456513 , -1.6530298 , -1.4865901 , -2.599905  ]]],
      dtype=float32), array([[[3.0081749e-02, 4.5505047e-02],
        [3.8255841e-02, 4.3677360e-02],
        [2.9920518e-02, 4.5374095e-02],
        ...,
        [9.5930613e-05, 1.5463531e-03],
        [1.2648106e-04, 1.9661188e-03],
        [1.8099880e-05, 5.3441525e-04]]], dtype=float32)]

        var y_bboxes = result[0]; //<--- Error
        var y_scores = result[1]; //<--- Error
        //Python.Runtime.PythonException: 'TypeError : list indices must be integers or slices, not tuple'
NDarray result=m_Model.Predict(image_exp);
[数组([[0.17123657,0.18411542,0.2141329,0.89467835],
[ 0.81485975,  1.0284375 , -0.41349524,  0.6665468 ],
[ 0.01628919,  0.11503349,  0.8445237 ,  0.06173617],
...,
[-1.1167231 , -1.4015226 , -2.944187  , -0.9941753 ],
[-1.4996854 , -1.6622332 , -2.0316055 , -1.8165188 ],
[-1.4456513 , -1.6530298 , -1.4865901 , -2.599905  ]]],
数据类型=float32),数组([[3.0081749e-02,4.5505047e-02],
[3.8255841e-02,4.3677360e-02],
[2.9920518e-02,4.5374095e-02],
...,
[9.5930613e-05,1.5463531e-03],
[1.2648106e-04,1.9661188e-03],
[1.8099880e-05,5.3441525e-04]],数据类型=32]
var y_bboxes=结果[0]//