I would like to implement ML-GCN using tensorflow and keras. Briefly, it is a two-head neural network: One head is a graph convolutional network (G
tensorflow
keras