Abstract
We present a new deep network layer called “Dynamic Convolutional Layer” which is a generalization of the convolutional layer. The conventional convolutional layer uses fifilters that are learned during training and are held constant during testing. In contrast, the dynamic convolutional layer uses fifilters that will vary from input to input during testing. This is achieved by learning a function that maps the input to the fifilters. We apply the dynamic convolutional layer to the application of short range weather prediction and show performance improvements compared to other baselines