Generative Models

This work proposes a novel approach to capture the uncertainty of the forecast using conditional Normalizing Flows, which allows us to model complex multimodal distributions and to compare various models using negative log-likelihood (NLL).

This work proposes a novel neural network architecture capable of continual learning on 3D point cloud data by using rehearsal and reconstruction as regularization methods of the learning process and achieves a significant decrease of catastrophic forgetting.