Patrick Takenaka
Publication
ViPro: Enabling and Controlling Video Prediction for Complex Dynamical Scenarios using Procedural Knowledge
We are happy to share that the work of our PhD student Patrick Takenaka was selected as a spotlight paper and was presented at the 18th International Conference on Neural-Symbolic Learning and Reasoning (NeSy 2024) in Barcelona, Spain.
Abstract
We propose a novel architecture design for video prediction in order to utilize procedural domain knowledge directly as part of the computational graph of data-driven models. On the basis of new challeng- ing scenarios we show that state-of-the-art video predictors struggle in complex dynamical settings, and highlight that the introduction of prior process knowledge makes their learning problem feasible. Our approach results in the learning of a symbolically addressable interface between data-driven aspects in the model and our dedicated procedural knowledge module, which we utilize in downstream control tasks.
Authors: Patrick Takenaka, Johannes Maucher, Marco F. Huber
Link to Paper (preprint): https://arxiv.org/abs/2407.09537
Link to Proceedings: https://link.springer.com/book/10.1007/978-3-031-71170-1
