In Keras, you would use a 1D convnet via the Conv1D layer. There exist also, Conv2D and Conv3D. It takes as input 3D tensors with shape (samples, time, ...
確定! 回上一頁