31. July 2021 - Providentia Editors

FloMo: Tractable Motion Prediction with Normalizing Flows

The future motion of traffic participants is inherently uncertain. To plan safely, therefore, an autonomous agent must take into account multiple possible trajectory outcomes and prioritize them. Recently, this problem has been addressed with generative neural networks. However, most generative models either do not learn the true underlying trajectory distribution reliably, or do not allow predictions to be associated with likelihoods. In our work, we model motion prediction directly as a density estimation problem with a normalizing flow between a noise distribution and the future motion distribution. Our model, named FloMo, allows likelihoods to be computed in a single network pass and can be trained directly with maximum likelihood estimation. Furthermore, we propose a method to stabilize training flows on trajectory datasets and a new data augmentation transformation that improves the performance and generalization of our model. Our method achieves state-of-the-art performance on three popular prediction datasets, with a significant gap to most competing models.




6. August 2021

Expanding the A9 test field: Improvisation required

Expanding the A9 test section for autonomous and connected driving required a great deal of research and pragmatic decision-making. Now that it has been accomplished and the scientists are at work, we look back at eight challenges and their solutions.


6. August 2021

Connected car services with great potential

Vehicle connectivity and automation is becoming increasingly important in the automotive industry. Particularly high growth potential is expected for connected car services, according to a recent study by IW Consult and Fraunhofer IAO.


31. July 2021

Predicting movements by autonomous agents

To predict movements, many trajectory options must be taken into consideration and prioritized. The FloMo model works with probabilities and has been trained on the basis of three datasets. Its performance is improved by a proposed method of stabilizing training flows.