1 code implementation • 16 Oct 2023 • Manuel Traub, Frederic Becker, Adrian Sauter, Sebastian Otte, Martin V. Butz
Current slot-oriented approaches for compositional scene segmentation from images and videos rely on provided background information or slot assignments.
no code implementations • 16 Oct 2023 • Manuel Traub, Frederic Becker, Sebastian Otte, Martin V. Butz
While human infants exhibit knowledge about object permanence from two months of age onwards, deep-learning approaches still largely fail to recognize objects' continued existence.
no code implementations • 11 Sep 2023 • Matthias Karlbauer, Nathaniel Cresswell-Clay, Raul A. Moreno, Dale R. Durran, Thorsten Kurth, Martin V. Butz
We present a parsimonious deep learning weather prediction model on the Hierarchical Equal Area isoLatitude Pixelization (HEALPix) to forecast seven atmospheric variables for arbitrarily long lead times on a global approximately 110 km mesh at 3h time resolution.
no code implementations • 6 Apr 2023 • Jannik Thuemmel, Matthias Karlbauer, Sebastian Otte, Christiane Zarfl, Georg Martius, Nicole Ludwig, Thomas Scholten, Ulrich Friedrich, Volker Wulfmeyer, Bedartha Goswami, Martin V. Butz
Deep learning has gained immense popularity in the Earth sciences as it enables us to formulate purely data-driven models of complex Earth system processes.
no code implementations • 18 Aug 2022 • Manfred Eppe, Christian Gumbsch, Matthias Kerzel, Phuong D. H. Nguyen, Martin V. Butz, Stefan Wermter
According to cognitive psychology and related disciplines, the development of complex problem-solving behaviour in biological agents depends on hierarchical cognitive mechanisms.
Hierarchical Reinforcement Learning reinforcement-learning +1
1 code implementation • 4 Jun 2022 • Christian Gumbsch, Maurits Adam, Birgit Elsner, Georg Martius, Martin V. Butz
Humans can make predictions on various time scales and hierarchical levels.
no code implementations • 1 Jun 2022 • Franziska Kaltenberger, Sebastian Otte, Martin V. Butz
Our system flexibly binds the information of the rotating figure into the alternative attractors resolving the illusion's ambiguity and imagining the respective depth interpretation and the corresponding direction of rotation.
1 code implementation • 26 May 2022 • Manuel Traub, Sebastian Otte, Tobias Menge, Matthias Karlbauer, Jannik Thümmel, Martin V. Butz
Moreover, it can anticipate object motion and interactions, which are crucial abilities for conceptual planning and reasoning.
Ranked #1 on Video Object Tracking on CATER
no code implementations • 23 Feb 2022 • Fedor Scholz, Christian Gumbsch, Sebastian Otte, Martin V. Butz
We show that our architecture, which is trained end-to-end to minimize an approximation of free energy, develops latent states that can be interpreted as affordance maps.
1 code implementation • 23 Nov 2021 • Matthias Karlbauer, Timothy Praditia, Sebastian Otte, Sergey Oladyshkin, Wolfgang Nowak, Martin V. Butz
We introduce a compositional physics-aware FInite volume Neural Network (FINN) for learning spatiotemporal advection-diffusion processes.
1 code implementation • NeurIPS 2021 • Christian Gumbsch, Martin V. Butz, Georg Martius
A common approach to prediction and planning in partially observable domains is to use recurrent neural networks (RNNs), which ideally develop and maintain a latent memory about hidden, task-relevant factors.
no code implementations • 12 May 2021 • Dania Humaidan, Sebastian Otte, Christian Gumbsch, Charley Wu, Martin V. Butz
A critical challenge for any intelligent system is to infer structure from continuous data streams.
1 code implementation • 13 Apr 2021 • Timothy Praditia, Matthias Karlbauer, Sebastian Otte, Sergey Oladyshkin, Martin V. Butz, Wolfgang Nowak
To tackle this issue, we introduce a new approach called the Finite Volume Neural Network (FINN).
no code implementations • ICLR Workshop Learning_to_Learn 2021 • Sarah Fabi, Sebastian Otte, Martin V. Butz
One aspect of learning to learn concerns the development of compositional knowledge structures that can be flexibly recombined in a semantically meaningful manner to analogically solve related problems.
no code implementations • 18 Dec 2020 • Manfred Eppe, Christian Gumbsch, Matthias Kerzel, Phuong D. H. Nguyen, Martin V. Butz, Stefan Wermter
We then relate these insights with contemporary hierarchical reinforcement learning methods, and identify the key machine intelligence approaches that realise these mechanisms.
Hierarchical Reinforcement Learning reinforcement-learning +1
no code implementations • 9 Dec 2020 • Mahdi Sadeghi, Fabian Schrodt, Sebastian Otte, Martin V. Butz
Evaluations show that the resulting gradient-based inference process solves the perspective taking and binding problem for known biological motion patterns, essentially yielding a Gestalt perception mechanism.
no code implementations • 2 Oct 2020 • Sebastian Otte, Matthias Karlbauer, Martin V. Butz
We introduce Active Tuning, a novel paradigm for optimizing the internal dynamics of recurrent neural networks (RNNs) on the fly.
no code implementations • 21 Sep 2020 • Matthias Karlbauer, Tobias Menge, Sebastian Otte, Hendrik P. A. Lensch, Thomas Scholten, Volker Wulfmeyer, Martin V. Butz
Knowledge about the hidden factors that determine particular system dynamics is crucial for both explaining them and pursuing goal-directed interventions.
no code implementations • 19 Sep 2020 • Matthias Karlbauer, Sebastian Otte, Hendrik P. A. Lensch, Thomas Scholten, Volker Wulfmeyer, Martin V. Butz
The novel DISTributed Artificial neural Network Architecture (DISTANA) is a generative, recurrent graph convolution neural network.
no code implementations • 12 May 2020 • Dania Humaidan, Sebastian Otte, Martin V. Butz
Here, we introduce a hierarchical, surprise-gated recurrent neural network architecture, which models this process and develops compact compressions of distinct event-like contexts.
no code implementations • 8 May 2020 • Manuel Traub, Martin V. Butz, R. Harald Baayen, Sebastian Otte
As a consequence, this limits in principle the inherent advantage of SNNs, that is, the potential to develop codes that rely on precise relative spike timings.
no code implementations • 16 Apr 2020 • Sarah Fabi, Sebastian Otte, Jonas Gregor Wiese, Martin V. Butz
In the past, the character challenge was only met by complex algorithms that were provided with stochastic primitives.
1 code implementation • 23 Dec 2019 • Matthias Karlbauer, Sebastian Otte, Hendrik P. A. Lensch, Thomas Scholten, Volker Wulfmeyer, Martin V. Butz
We introduce a distributed spatio-temporal artificial neural network architecture (DISTANA).
no code implementations • 26 Feb 2019 • Christian Gumbsch, Martin V. Butz, Georg Martius
Here, we introduce a computational learning architecture, termed surprise-based behavioral modularization into event-predictive structures (SUBMODES), that explores behavior and identifies the underlying behavioral units completely from scratch.
2 code implementations • 19 Sep 2018 • Martin V. Butz, David Bilkey, Dania Humaidan, Alistair Knott, Sebastian Otte
We introduce REPRISE, a REtrospective and PRospective Inference SchEme, which learns temporal event-predictive models of dynamical systems.