Search Results for author:
Found 78 papers, 50 papers with code
DeSplat: Decomposed Gaussian Splatting for Distractor-Free Rendering
Gaussian splatting enables fast novel view synthesis in static 3D environments.
Streamlining Prediction in Bayesian Deep Learning
The rising interest in Bayesian deep learning (BDL) has led to a plethora of methods for estimating the posterior distribution.
Differentially Private Continual Learning using Pre-Trained Models
This work explores the intersection of continual learning (CL) and differential privacy (DP).
Plan$\times$RAG: Planning-guided Retrieval Augmented Generation
We introduce Planning-guided Retrieval Augmented Generation (Plan$\times$RAG), a novel framework that augments the \emph{retrieve-then-reason} paradigm of existing RAG frameworks to \emph{plan-then-retrieve}.
Free Hunch: Denoiser Covariance Estimation for Diffusion Models Without Extra Costs
The covariance for clean data given a noisy observation is an important quantity in many conditional generation methods for diffusion models.
Physics-Informed Variational State-Space Gaussian Processes
Differential equations are important mechanistic models that are integral to many scientific and engineering applications.
Sources of Uncertainty in 3D Scene Reconstruction
The process of 3D scene reconstruction can be affected by numerous uncertainty sources in real-world scenes.
Exploiting Hankel-Toeplitz Structures for Fast Computation of Kernel Precision Matrices
In this paper, we lower this dominating computational complexity to $\mathcal{O}(NM)$ with no additional approximations.
iQRL -- Implicitly Quantized Representations for Sample-efficient Reinforcement Learning
Learning representations for reinforcement learning (RL) has shown much promise for continuous control.
Improving Discrete Diffusion Models via Structured Preferential Generation
In the domains of image and audio, diffusion models have shown impressive performance.
Alignment is Key for Applying Diffusion Models to Retrosynthesis
Retrosynthesis, the task of identifying precursors for a given molecule, can be naturally framed as a conditional graph generation task.
Flatness Improves Backbone Generalisation in Few-shot Classification
Deployment of deep neural networks in real-world settings typically requires adaptation to new tasks with few examples.
Gaussian Splatting on the Move: Blur and Rolling Shutter Compensation for Natural Camera Motion
High-quality scene reconstruction and novel view synthesis based on Gaussian Splatting (3DGS) typically require steady, high-quality photographs, often impractical to capture with handheld cameras.
Function-space Parameterization of Neural Networks for Sequential Learning
Our parameterization offers: (i) a way to scale function-space methods to large data sets via sparsification, (ii) retention of prior knowledge when access to past data is limited, and (iii) a mechanism to incorporate new data without retraining.
Subtractive Mixture Models via Squaring: Representation and Learning
Mixture models are traditionally represented and learned by adding several distributions as components.
The Robust Semantic Segmentation UNCV2023 Challenge Results
This paper outlines the winning solutions employed in addressing the MUAD uncertainty quantification challenge held at ICCV 2023.
Sparse Function-space Representation of Neural Networks
Deep neural networks (NNs) are known to lack uncertainty estimates and struggle to incorporate new data.
Improving Hyperparameter Learning under Approximate Inference in Gaussian Process Models
Approximate inference in Gaussian process (GP) models with non-conjugate likelihoods gets entangled with the learning of the model hyperparameters.
Memory-Based Dual Gaussian Processes for Sequential Learning
Sequential learning with Gaussian processes (GPs) is challenging when access to past data is limited, for example, in continual and active learning.
Variational Gaussian Process Diffusion Processes
Diffusion processes are a class of stochastic differential equations (SDEs) providing a rich family of expressive models that arise naturally in dynamic modelling tasks.
PriorCVAE: scalable MCMC parameter inference with Bayesian deep generative modelling
Recent advances have shown that GP priors, or their finite realisations, can be encoded using deep generative models such as variational autoencoders (VAEs).
Fixing Overconfidence in Dynamic Neural Networks
Dynamic neural networks are a recent technique that promises a remedy for the increasing size of modern deep learning models by dynamically adapting their computational cost to the difficulty of the inputs.
Transport with Support: Data-Conditional Diffusion Bridges
We integrate Bayesian filtering and optimal control into learning the diffusion process, enabling the generation of constrained stochastic processes governed by sparse observations at intermediate stages and terminal constraints.
MixupE: Understanding and Improving Mixup from Directional Derivative Perspective
Mixup is a popular data augmentation technique for training deep neural networks where additional samples are generated by linearly interpolating pairs of inputs and their labels.
Towards Improved Learning in Gaussian Processes: The Best of Two Worlds
Gaussian process training decomposes into inference of the (approximate) posterior and learning of the hyperparameters.
Fantasizing with Dual GPs in Bayesian Optimization and Active Learning
Gaussian processes (GPs) are the main surrogate functions used for sequential modelling such as Bayesian Optimization and Active Learning.
Expansion of Visual Hints for Improved Generalization in Stereo Matching
We introduce visual hints expansion for guiding stereo matching to improve generalization.
Uncertainty-guided Source-free Domain Adaptation
Source-free domain adaptation (SFDA) aims to adapt a classifier to an unlabelled target data set by only using a pre-trained source model.
Generative Modelling With Inverse Heat Dissipation
While diffusion models have shown great success in image generation, their noise-inverting generative process does not explicitly consider the structure of images, such as their inherent multi-scale nature.
Disentangling Model Multiplicity in Deep Learning
This observed 'predictive' multiplicity (PM) also implies elusive differences in the internals of the models, their 'representational' multiplicity (RM).
A Look at Improving Robustness in Visual-inertial SLAM by Moment Matching
The fusion of camera sensor and inertial data is a leading method for ego-motion tracking in autonomous and smart devices.
Non-separable Spatio-temporal Graph Kernels via SPDEs
Gaussian processes (GPs) provide a principled and direct approach for inference and learning on graphs.
Dual Parameterization of Sparse Variational Gaussian Processes
Sparse variational Gaussian process (SVGP) methods are a common choice for non-conjugate Gaussian process inference because of their computational benefits.
Bayes-Newton Methods for Approximate Bayesian Inference with PSD Guarantees
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.
Spatio-Temporal Variational Gaussian Processes
We introduce a scalable approach to Gaussian process inference that combines spatio-temporal filtering with natural gradient variational inference, resulting in a non-conjugate GP method for multivariate data that scales linearly with respect to time.
Scalable Inference in SDEs by Direct Matching of the Fokker-Planck-Kolmogorov Equation
Simulation-based techniques such as variants of stochastic Runge-Kutta are the de facto approach for inference with stochastic differential equations (SDEs) in machine learning.
Periodic Activation Functions Induce Stationarity
Neural network models are known to reinforce hidden data biases, making them unreliable and difficult to interpret.
Sparse Gaussian Processes for Stochastic Differential Equations
We frame the problem of learning stochastic differential equations (SDEs) from noisy observations as an inference problem and aim to maximize the marginal likelihood of the observations in a joint model of the latent paths and the noisy observations.
HybVIO: Pushing the Limits of Real-time Visual-inertial Odometry
We present HybVIO, a novel hybrid approach for combining filtering-based visual-inertial odometry (VIO) with optimization-based SLAM.
Combining Pseudo-Point and State Space Approximations for Sum-Separable Gaussian Processes
Pseudo-point approximations, one of the gold-standard methods for scaling GPs to large data sets, are well suited for handling off-the-grid spatial data.
Scalable Inference in SDEs by Direct Matching of the Fokker–Planck–Kolmogorov Equation
Simulation-based techniques such as variants of stochastic Runge–Kutta are the de facto approach for inference with stochastic differential equations (SDEs) in machine learning.
Sparse Algorithms for Markovian Gaussian Processes
Approximate Bayesian inference methods that scale to very large datasets are crucial in leveraging probabilistic models for real-world time series.
Novel View Synthesis via Depth-guided Skip Connections
Flow predictions enable the target view to re-use pixels directly, but can easily lead to distorted results.
RealAnt: An Open-Source Low-Cost Quadruped for Education and Research in Real-World Reinforcement Learning
We validate the platform with reinforcement learning experiments and provide baseline results on a set of benchmark tasks.
Stationary Activations for Uncertainty Calibration in Deep Learning
We introduce a new family of non-linear neural network activation functions that mimic the properties induced by the widely-used Mat\'ern family of kernels in Gaussian process (GP) models.
Movement-induced Priors for Deep Stereo
We propose a method for fusing stereo disparity estimation with movement-induced prior information.
State Space Expectation Propagation: Efficient Inference Schemes for Temporal Gaussian Processes
EP provides some benefits over the traditional methods via introduction of the so-called cavity distribution, and we combine these benefits with the computational efficiency of linearisation, providing extensive empirical analysis demonstrating the efficacy of various algorithms under this unifying framework.
Fast Variational Learning in State-Space Gaussian Process Models
Gaussian process (GP) regression with 1D inputs can often be performed in linear time via a stochastic differential equation formulation.
Movement Tracking by Optical Flow Assisted Inertial Navigation
We consider a visually dense approach, where the IMU data is fused with the dense optical flow field estimated from the camera data.
Deep Residual Mixture Models
We propose Deep Residual Mixture Models (DRMMs), a novel deep generative model architecture.
Deep Automodulators
These networks can faithfully reproduce individual real-world input images like regular autoencoders, but also generate a fused sample from an arbitrary combination of several such images, allowing instantaneous 'style-mixing' and other new applications.
Gaussian Process Priors for View-Aware Inference
While frame-independent predictions with deep neural networks have become the prominent solutions to many computer vision tasks, the potential benefits of utilizing correlations between frames have received less attention.
Scalable Exact Inference in Multi-Output Gaussian Processes
Multi-output Gaussian processes (MOGPs) leverage the flexibility and interpretability of GPs while capturing structure across outputs, which is desirable, for example, in spatio-temporal modelling.
Global Approximate Inference via Local Linearisation for Temporal Gaussian Processes
The extended Kalman filter (EKF) is a classical signal processing algorithm which performs efficient approximate Bayesian inference in non-conjugate models by linearising the local measurement function, avoiding the need to compute intractable integrals when calculating the posterior.
Iterative Path Reconstruction for Large-Scale Inertial Navigation on Smartphones
Modern smartphones have all the sensing capabilities required for accurate and robust navigation and tracking.
Towards Photographic Image Manipulation with Balanced Growing of Generative Autoencoders
retaining the identity of a face), sharp generated/reconstructed samples in high resolutions, and a well-structured latent space that supports semantic manipulation of the inputs.
Multi-View Stereo by Temporal Nonparametric Fusion
The flexibility of the Gaussian process (GP) prior provides adapting memory for fusing information from previous views.
Know Your Boundaries: Constraining Gaussian Processes by Variational Harmonic Features
Gaussian processes (GPs) provide a powerful framework for extrapolation, interpolation, and noise removal in regression and classification.
Interpolation Consistency Training for Semi-Supervised Learning
We introduce Interpolation Consistency Training (ICT), a simple and computation efficient algorithm for training Deep Neural Networks in the semi-supervised learning paradigm.
Unstructured Multi-View Depth Estimation Using Mask-Based Multiplane Representation
This paper presents a novel method, MaskMVS, to solve depth estimation for unstructured multi-view image-pose pairs.
End-to-End Probabilistic Inference for Nonstationary Audio Analysis
A typical audio signal processing pipeline includes multiple disjoint analysis stages, including calculation of a time-frequency representation followed by spectrogram-based feature analysis.
Infinite-Horizon Gaussian Processes
The complexity is still cubic in the state dimension $m$ which is an impediment to practical application.
Unifying Probabilistic Models for Time-Frequency Analysis
In audio signal processing, probabilistic time-frequency models have many benefits over their non-probabilistic counterparts.
Deep Learning Based Speed Estimation for Constraining Strapdown Inertial Navigation on Smartphones
Strapdown inertial navigation systems are sensitive to the quality of the data provided by the accelerometer and gyroscope.
ADVIO: An authentic dataset for visual-inertial odometry
The lack of realistic and open benchmarking datasets for pedestrian visual-inertial odometry has made it hard to pinpoint differences in published methods.
Pioneer Networks: Progressively Growing Generative Autoencoder
Instead, we propose the Progressively Growing Generative Autoencoder (PIONEER) network which achieves high-quality reconstruction with $128{\times}128$ images without requiring a GAN discriminator.
Robust Gyroscope-Aided Camera Self-Calibration
This application paper proposes a model for estimating the parameters on the fly by fusing gyroscope and camera data, both readily available in modern day smartphones.
Scalable Magnetic Field SLAM in 3D Using Gaussian Process Maps
We present a method for scalable and fully 3D magnetic field simultaneous localisation and mapping (SLAM) using local anomalies in the magnetic field as a source of position information.
Recursive Chaining of Reversible Image-to-image Translators For Face Aging
By treating the age phases as a sequence of image domains, we construct a chain of transformers that map images from one age domain to the next.
State Space Gaussian Processes with Non-Gaussian Likelihood
We provide a comprehensive overview and tooling for GP modeling with non-Gaussian likelihoods using state space methods.
PIVO: Probabilistic Inertial-Visual Odometry for Occlusion-Robust Navigation
This paper presents a novel method for visual-inertial odometry.
Inertial Odometry on Handheld Smartphones
Building a complete inertial navigation system using the limited quality data provided by current smartphones has been regarded challenging, if not impossible.
Variational Fourier features for Gaussian processes
This work brings together two powerful concepts in Gaussian processes: the variational approach to sparse approximation and the spectral representation of Gaussian processes.
Regularizing Solutions to the MEG Inverse Problem Using Space-Time Separable Covariance Functions
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.
Modeling and interpolation of the ambient magnetic field by Gaussian processes
Anomalies in the ambient magnetic field can be used as features in indoor positioning and navigation.
Computationally Efficient Bayesian Learning of Gaussian Process State Space Models
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.
Sigma-Point Filtering and Smoothing Based Parameter Estimation in Nonlinear Dynamic Systems
We consider approximate maximum likelihood parameter estimation in nonlinear state-space models.
Methodology Dynamical Systems Optimization and Control Computation
Hilbert Space Methods for Reduced-Rank Gaussian Process Regression
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.
