no code implementations • 22 Mar 2024 • Nutan Chen, Elie Aljalbout, Botond Cseke, Patrick van der Smagt
This integration facilitates rapid adaptation to new tasks and optimizes the utilization of accumulated expertise by allowing robots to learn and generalize from demonstrated trajectories.
no code implementations • 13 Jun 2022 • Nutan Chen, Patrick van der Smagt, Botond Cseke
Auto-encoder models that preserve similarities in the data are a popular tool in representation learning.
no code implementations • NeurIPS 2021 • Alexej Klushyn, Richard Kurle, Maximilian Soelch, Botond Cseke, Patrick van der Smagt
Our results show that the constrained optimisation framework significantly improves system identification and prediction accuracy on the example of established state-of-the-art DSSMs.
no code implementations • 29 Jan 2021 • Felix Frank, Alexandros Paraschos, Patrick van der Smagt, Botond Cseke
We unify previous adaptation techniques, for example, various types of obstacle avoidance, via-points, mutual avoidance, in one single framework and combine them to solve complex robotic problems.
Robotics
no code implementations • ICLR 2020 • Richard Kurle, Botond Cseke, Alexej Klushyn, Patrick van der Smagt, Stephan Günnemann
We represent the posterior approximation of the network weights by a diagonal Gaussian distribution and a complementary memory of raw data.
no code implementations • 23 Aug 2019 • Alexej Klushyn, Nutan Chen, Botond Cseke, Justin Bayer, Patrick van der Smagt
We address the problem of one-to-many mappings in supervised learning, where a single instance has many different solutions of possibly equal cost.
no code implementations • NeurIPS 2019 • Alexej Klushyn, Nutan Chen, Richard Kurle, Botond Cseke, Patrick van der Smagt
We propose to learn a hierarchical prior in the context of variational autoencoders to avoid the over-regularisation resulting from a standard normal prior distribution.
no code implementations • 1 Jun 2017 • David Schnoerr, Botond Cseke, Ramon Grima, Guido Sanguinetti
We consider the problem of computing first-passage time distributions for reaction processes modelled by master equations.
2 code implementations • NeurIPS 2016 • Sebastian Nowozin, Botond Cseke, Ryota Tomioka
Generative neural samplers are probabilistic models that implement sampling using feedforward neural networks: they take a random input vector and produce a sample from a probability distribution defined by the network weights.
no code implementations • 18 Dec 2015 • Botond Cseke, David Schnoerr, Manfred Opper, Guido Sanguinetti
We consider the inverse problem of reconstructing the posterior measure over the trajec- tories of a diffusion process from discrete time observations and continuous time constraints.
no code implementations • 16 Jan 2014 • Botond Cseke, Tom Heskes
We define the Gaussian fractional Bethe free energy in terms of the moment parameters of the approximate marginals, derive a lower and an upper bound on the fractional Bethe free energy and establish a necessary condition for the lower bound to be bounded from below.
no code implementations • NeurIPS 2013 • Botond Cseke, Manfred Opper, Guido Sanguinetti
We propose an approximate inference algorithm for continuous time Gaussian-Markov process models with both discrete and continuous time likelihoods.
no code implementations • 17 May 2013 • Botond Cseke, Guido Sanguinetti
We consider the problem of joint modelling of metabolic signals and gene expression in systems biology applications.
no code implementations • 17 May 2013 • Botond Cseke, Andrew Zammit Mangion, Tom Heskes, Guido Sanguinetti
Spatio-temporal point process models play a central role in the analysis of spatially distributed systems in several disciplines.
no code implementations • NeurIPS 2009 • Marcel V. Gerven, Botond Cseke, Robert Oostenveld, Tom Heskes
We introduce a novel multivariate Laplace (MVL) distribution as a sparsity promoting prior for Bayesian source localization that allows the specification of constraints between and within sources.