no code implementations • 27 Feb 2024 • Harshit Agrawal, Ari Hietanen, Simo Särkkä
Our findings suggest that this novel approach outperforms the baseline U-Net, offering a significant step towards practical application in real clinical settings where CBCT systems are employed to scan a wide range of FOVs.
no code implementations • 12 Feb 2024 • Sahel Iqbal, Adrien Corenflos, Simo Särkkä, Hany Abdulsamad
In this paper, we propose a novel approach to Bayesian experimental design for non-exchangeable data that formulates it as risk-sensitive policy optimization.
no code implementations • 1 Feb 2024 • Ahmad Farooq, Cristian A. Galvis-Florez, Simo Särkkä
Gaussian processes are probabilistic models that are commonly used as functional priors in machine learning.
1 code implementation • 21 Dec 2023 • Hany Abdulsamad, Sahel Iqbal, Adrien Corenflos, Simo Särkkä
Stochastic optimal control of dynamical systems is a crucial challenge in sequential decision-making.
1 code implementation • 6 Nov 2023 • Arina Odnoblyudova, Çağlar Hızlı, ST John, Andrea Cognolato, Anne Juuti, Simo Särkkä, Kirsi Pietiläinen, Pekka Marttinen
By differentiating treatment components, incorporating their dosages, and sharing statistical information across patients via a hierarchical multi-output Gaussian process, our method improves prediction accuracy over existing approaches, and allows us to interpret the different effects of carbohydrates and fat on the overall glucose response.
1 code implementation • 2 Oct 2023 • Nathanael Bosch, Adrien Corenflos, Fatemeh Yaghoobi, Filip Tronarp, Philipp Hennig, Simo Särkkä
Probabilistic numerical solvers for ordinary differential equations (ODEs) treat the numerical simulation of dynamical systems as problems of Bayesian state estimation.
no code implementations • 15 Jun 2023 • Fatemeh Yaghoobi, Hany Abdulsamad, Simo Särkkä
In this paper, we use the optimization formulation of nonlinear Kalman filtering and smoothing problems to develop second-order variants of iterated Kalman smoother (IKS) methods.
1 code implementation • 1 Mar 2023 • Adrien Corenflos, Simo Särkkä
We introduce two new classes of exact Markov chain Monte Carlo (MCMC) samplers for inference in latent dynamical models.
no code implementations • 17 Aug 2022 • Harshit Agrawal, Ari Hietanen, Simo Särkkä
We compare our model's performance on real clinical scans with conventional region growing threshold-based MAR, moving metal artifact reduction method, and a recent deep learning method.
1 code implementation • 15 Aug 2022 • Hao Dong, Xieyuanli Chen, Simo Särkkä, Cyrill Stachniss
We further use the extracted poles as pseudo labels to train a deep neural network for online range image-based pole segmentation.
1 code implementation • 12 May 2022 • Zheng Zhao, Simo Särkkä, Jens Sjölund, Thomas B. Schön
We present a continuous-time probabilistic approach for estimating the chirp signal and its instantaneous frequency function when the true forms of these functions are not accessible.
1 code implementation • 4 Feb 2022 • Adrien Corenflos, Nicolas Chopin, Simo Särkkä
We propose dSMC (de-Sequentialized Monte Carlo), a new particle smoother that is able to process $T$ observations in $\mathcal{O}(\log T)$ time on parallel architecture.
no code implementations • 22 Jan 2022 • Joel Jaskari, Jaakko Sahlsten, Theodoros Damoulas, Jeremias Knoblauch, Simo Särkkä, Leo Kärkkäinen, Kustaa Hietala, Kimmo Kaski
Automatic classification of diabetic retinopathy from retinal images has been widely studied using deep neural networks with impressive results.
1 code implementation • 2 Nov 2021 • William J. Wilkinson, Simo Särkkä, Arno Solin
We formulate natural gradient variational inference (VI), expectation propagation (EP), and posterior linearisation (PL) as extensions of Newton's method for optimising the parameters of a Bayesian posterior distribution.
no code implementations • 20 May 2021 • Zheng Zhao, Rui Gao, Simo Särkkä
This paper is concerned with regularized extensions of hierarchical non-stationary temporal Gaussian processes (NSGPs) in which the parameters (e. g., length-scale) are modeled as GPs.
no code implementations • 25 Apr 2021 • Christos Merkatas, Simo Särkkä
System identification is of special interest in science and engineering.
1 code implementation • 19 Feb 2021 • Adrien Corenflos, Zheng Zhao, Simo Särkkä
The aim of this article is to present a novel parallelization method for temporal Gaussian process (GP) regression problems.
1 code implementation • 10 Feb 2021 • Sakira Hassan, Simo Särkkä, Ángel F. García-Fernández
This paper presents algorithms for parallelization of inference in hidden Markov models (HMMs).
no code implementations • 1 Feb 2021 • Toni Karvonen, Jon Cockayne, Filip Tronarp, Simo Särkkä
We study a class of Gaussian processes for which the posterior mean, for a particular choice of data, replicates a truncated Taylor expansion of any order.
1 code implementation • 11 Aug 2020 • Zheng Zhao, Muhammad Emzir, Simo Särkkä
This paper is concerned with a state-space approach to deep Gaussian process (DGP) regression.
1 code implementation • 28 Jun 2020 • Muhammad Emzir, Sari Lasanen, Zenith Purisha, Lassi Roininen, Simo Särkkä
In this article, we study Bayesian inverse problems with multi-layered Gaussian priors.
Statistics Theory Statistics Theory
no code implementations • 17 Apr 2020 • Filip Tronarp, Simo Särkkä
For approximate filtering and smoothing the projection approach is taken, where it turns out that the prediction and smoothing equations are the same as in the case when the state variable evolves in Euclidean space.
no code implementations • 29 Jan 2020 • Toni Karvonen, George Wynne, Filip Tronarp, Chris. J. Oates, Simo Särkkä
We show that the maximum likelihood estimation of the scale parameter alone provides significant adaptation against misspecification of the Gaussian process model in the sense that the model can become "slowly" overconfident at worst, regardless of the difference between the smoothness of the data-generating function and that expected by the model.
2 code implementations • 8 Jan 2020 • Zheng Zhao, Toni Karvonen, Roland Hostettler, Simo Särkkä
The paper is concerned with non-linear Gaussian filtering and smoothing in continuous-discrete state-space models, where the dynamic model is formulated as an It\^{o} stochastic differential equation (SDE), and the measurements are obtained at discrete time instants.
no code implementations • 6 Sep 2019 • Ali Bahrami Rad, Morteza Zabihi, Zheng Zhao, Moncef Gabbouj, Aggelos K. Katsaggelos, Simo Särkkä
Results: The proposed algorithm is validated on the 2018 PhysioNet challenge dataset.
no code implementations • 13 Jul 2019 • Simo Särkkä
Gaussian process state-space models (GPSS) can be used to learn the dynamic and measurement models for a state-space representation of the input-output data.
no code implementations • 1 Mar 2019 • Morteza Zabihi, Ali Bahrami Rad, Serkan Kiranyaz, Simo Särkkä, Moncef Gabbouj
Sleep arousals transition the depth of sleep to a more superficial stage.
no code implementations • 12 Dec 2018 • Zheng Zhao, Simo Särkkä, Ali Bahrami Rad
In this article, we propose a novel ECG classification framework for atrial fibrillation (AF) detection using spectro-temporal representation (i. e., time varying spectrum) and deep convolutional networks.
no code implementations • 28 Nov 2018 • Jakub Prüher, Toni Karvonen, Chris. J. Oates, Ondřej Straka, Simo Särkkä
The sigma-point filters, such as the UKF, which exploit numerical quadrature to obtain an additional order of accuracy in the moment transformation step, are popular alternatives to the ubiquitous EKF.
no code implementations • 23 Nov 2018 • Janne Mustaniemi, Juho Kannala, Jiri Matas, Simo Särkkä, Janne Heikkilä
The paper addresses the problem of acquiring high-quality photographs with handheld smartphone cameras in low-light imaging conditions.
1 code implementation • 8 Oct 2018 • Filip Tronarp, Hans Kersting, Simo Särkkä, Philipp Hennig
We formulate probabilistic numerical approximations to solutions of ordinary differential equations (ODEs) as problems in Gaussian process (GP) regression with non-linear measurement functions.
1 code implementation • 1 Oct 2018 • Janne Mustaniemi, Juho Kannala, Simo Särkkä, Jiri Matas, Janne Heikkilä
We propose a deblurring method that incorporates gyroscope measurements into a convolutional neural network (CNN).
1 code implementation • 26 Sep 2018 • Toni Karvonen, Simo Särkkä, Chris. J. Oates
Bayesian cubature provides a flexible framework for numerical integration, in which a priori knowledge on the integrand can be encoded and exploited.
Methodology Numerical Analysis Computation
no code implementations • 13 Sep 2018 • Ángel F. García-Fernández, Filip Tronarp, Simo Särkkä
This paper proposes a new algorithm for Gaussian process classification based on posterior linearisation (PL).
no code implementations • 11 Sep 2018 • Zenith Purisha, Carl Jidling, Niklas Wahlström, Simo Särkkä, Thomas B. Schön
The approach also allows for reformulation of come classical regularization methods as Laplacian and Tikhonov regularization as Gaussian process regression, and hence provides an efficient algorithm and principled means for their parameter tuning.
no code implementations • 22 May 2018 • Janne Mustaniemi, Juho Kannala, Simo Särkkä, Jiri Matas, Janne Heikkilä
It is well-known that motion blur decreases the performance of traditional feature detectors and descriptors.
no code implementations • 15 Sep 2017 • Simo Särkkä, Mauricio A. Álvarez, Neil D. Lawrence
This article is concerned with learning and stochastic control in physical systems which contain unknown input signals.
no code implementations • 15 Mar 2017 • Jakub Prüher, Filip Tronarp, Toni Karvonen, Simo Särkkä, Ondřej Straka
Advantage of the Student- t process quadrature over the traditional Gaussian process quadrature, is that the integral variance depends also on the function values, allowing for a more robust modelling of the integration error.
1 code implementation • 29 Nov 2016 • Janne Mustaniemi, Juho Kannala, Simo Särkkä, Jiri Matas, Janne Heikkilä
In the process, we also perform a temporal and spatial alignment of the camera and the IMU.
no code implementations • 25 Oct 2016 • Alexander Grigorievskiy, Neil Lawrence, Simo Särkkä
We propose a parallelizable sparse inverse formulation Gaussian process (SpInGP) for temporal models.
1 code implementation • 17 Oct 2016 • Michael Schober, Simo Särkkä, Philipp Hennig
Like many numerical methods, solvers for initial value problems (IVPs) on ordinary differential equations estimate an analytically intractable quantity, using the results of tractable computations as inputs.
no code implementations • 17 Apr 2016 • Arno Solin, Pasi Jylänki, Jaakko Kauramäki, Tom Heskes, Marcel A. J. van Gerven, Simo Särkkä
We apply the method to both simulated and empirical data, and demonstrate the efficiency and generality of our Bayesian source reconstruction approach which subsumes various classical approaches in the literature.
no code implementations • 15 Sep 2015 • Arno Solin, Manon Kok, Niklas Wahlström, Thomas B. Schön, Simo Särkkä
Anomalies in the ambient magnetic field can be used as features in indoor positioning and navigation.
no code implementations • 7 Jun 2015 • Andreas Svensson, Arno Solin, Simo Särkkä, Thomas B. Schön
We present a procedure for efficient Bayesian learning in Gaussian process state space models, where the representation is formed by projecting the problem onto a set of approximate eigenfunctions derived from the prior covariance structure.
1 code implementation • 23 Apr 2015 • Juho Kokkala, Arno Solin, Simo Särkkä
We consider approximate maximum likelihood parameter estimation in nonlinear state-space models.
Methodology Dynamical Systems Optimization and Control Computation
no code implementations • 22 Apr 2015 • Simo Särkkä, Jouni Hartikainen, Lennart Svensson, Fredrik Sandblom
This article is concerned with Gaussian process quadratures, which are numerical integration methods based on Gaussian process regression methods, and sigma-point methods, which are used in advanced non-linear Kalman filtering and smoothing algorithms.
2 code implementations • 21 Jan 2014 • Arno Solin, Simo Särkkä
On this approximate eigenbasis the eigenvalues of the covariance function can be expressed as simple functions of the spectral density of the Gaussian process, which allows the GP inference to be solved under a computational cost scaling as $\mathcal{O}(nm^2)$ (initial) and $\mathcal{O}(m^3)$ (hyperparameter learning) with $m$ basis functions and $n$ data points.
no code implementations • NeurIPS 2012 • Simon Lyons, Amos J. Storkey, Simo Särkkä
The decomposition allows us to take a diffusion process of interest and cast it in a form that is amenable to sampling by MCMC methods.