Trying to create a model for recognising sign language. Using 3d convolution layers and finally the LTC cell from NCP library. But the model accuracy does not improves after
