4 ms·[1901.11390] MONet: Unsupervised Scene Decomposition and Representation3 points by pplonski86 8y agorsaha 8y agoIs the second equation in the paper correct, shouldn't the mask be just the output of the attention network?