python - Pass Input to tensorflow lite model in Android

Question

I have created a neural network that take numerical data as input and saved it as tensorflow lite model using python. I am trying to pass input to the model in Android. Shape of ndarray is 1*3

Sample of the input in python is as follows

np.array([[-0.276786765 ,8.41897583008  ,-0.0222015380859]])

But i do not know to create the same input in java to pass it to model.

I tried using nd4j library. But still not able to write the proper code to create the input which is required by the model.

标签： pythontensorflowneural-networktensorflow-lite

Answer 1

假设您使用的是TensorFlow Lite，您可以使用以下方法提供 1x3 输入：

float[] innerInput = {-0.276786765 ,8.41897583008  ,-0.0222015380859
float[][] input = {innerInput};
interpreter.run(input, output);