Conference Papers (KKY)
Permanent URI for this collection
Browse
Recent Submissions
Item Neural Augmented Adaptive Grid Design for Point-Mass Filter(IEEE, 2025) Trejbal, Jan; Matoušek, Jakub; Duník, JindřichThis paper deals with the state estimation of nonlinear systems described by dynamic stochastic state-space models using a point-mass filter (PMF). The PMF is based on the approximation of the conditional probability density function by a piece-wise constant probability density, called the point-mass density (PMD), where the probability is evaluated at N grid points. The number of grid points significantly affects both the performance and computational complexity of the PMF. However, N is typically regarded as a user-defined parameter. The aim of this paper is to augment the PMF with a neural network (NN). This NN selects the smallest N that leads to the required estimation accuracy thus ensuring the minimal computational complexity.
Item Diffusion in Lagrangian Grid-based Predictors(IEEE, 2025) Matoušek, Jakub; Duník, Jindřich; Govaers, Felix; Gehlen, JoshuaThis paper focuses on state prediction for stochastic dynamic models with linear dynamics, emphasizing a recently proposed efficient and robust Lagrangian approach for solving the Chapman-Kolmogorov equation. In contrast to the standard Eulerian perspective, the Lagrangian method separates the solution into two sequential steps: advection and diffusion. Advection is handled by moving a carefully designed grid, while diffusion is addressed using the convolution theorem. This approach significantly reduces computational complexity while preserving the same accuracy. In this paper, we propose formulating diffusion as a continuous-time process, leading to a partial differential equation (PDE). Various methods for solving this PDE are presented and compared within a unified framework, along with evaluations of their properties and example implementations. We demonstrate that the continuous formulation can yield substantial reductions in computational complexity with only marginal loss in accuracy.Item The Right Tracker for the Right Job: Maritime Case Study(Západočeská univerzita v Plzni, 2025) Krejčí, JanMulti-object tracking refers to maintaining awareness of the positions/velocities of multiple moving objects. It is a key technology for naval or airborne applications, autonomous driving, or space situational awareness, among which many are safety-critical. A large number of tracking algorithms, i.e., trackers, have been developed over the past decades, each having different complexity and performance. The distinction between various trackers is seldom clear. To allow for automatic tracker selection, a user-adjustable performance evaluation metric is needed. This contribution extends the discussion by exploring the trajectory generalized optimal sub-pattern assignment (TGOSPA) metric in a naval tracking example.Item The Right Tracker for the Right Job: Maritime Case Study(Západočeská univerzita v Plzni, 2025) Krejčí, JanMulti-object tracking refers to maintaining awareness of the positions/velocities of multiple moving objects. It is a key technology for naval or airborne applications, autonomous driving, or space situational awareness, among which many are safety-critical. A large number of tracking algorithms, i.e., trackers, have been developed over the past decades, each having different complexity and performance. The distinction between various trackers is seldom clear. To allow for automatic tracker selection, a user-adjustable performance evaluation metric is needed. This contribution extends the discussion by exploring the trajectory generalized optimal sub-pattern assignment (TGOSPA) metric in a naval tracking example.Item The Right Tracker for the Right Job: Maritime Case Study(Západočeská univerzita v Plzni, 2025) Krejčí, JanMulti-object tracking refers to maintaining awareness of the positions/velocities of multiple moving objects. It is a key technology for naval or airborne applications, autonomous driving, or space situational awareness, among which many are safety-critical. A large number of tracking algorithms, i.e., trackers, have been developed over the past decades, each having different complexity and performance. The distinction between various trackers is seldom clear. To allow for automatic tracker selection, a user-adjustable performance evaluation metric is needed. This contribution extends the discussion by exploring the trajectory generalized optimal sub-pattern assignment (TGOSPA) metric in a naval tracking example.
Item Model-Based Multi-Object Visual Tracking: Identification and Standard Model Limitations(IEEE, 2025) Krejčí, Jan; Kost, Oliver; Xia, Yuxuan; Svensson, Lennart; Straka, OndřejThis paper uses multi-object tracking methods known from the radar tracking community to address the problem of pedestrian tracking using 2D bounding box detections. The standard point-object (SPO) model is adopted, and the posterior density is computed using the Poisson multi-Bernoulli mixture (PMBM) filter. The selection of the model parameters rooted in continuous time is discussed, including the birth and survival probabilities. Some parameters are selected from the first principles, while others are identified from the data, which is, in this case, the publicly available MOT-17 dataset. Although the resulting PMBM algorithm yields promising results, a mismatch between the SPO model and the data is revealed. The model-based approach assumes that modifying the problematic components causing the SPO model-data mismatch will lead to better modelbased algorithms in future developments.Item Point-mass Filter with Non-equidistant Grid Design(Západočeská univerzita v Plzni, 2025) Trejbal, JanState estimation is essential in engineering applications from navigation to control. While many estimation methods assume Gaussian probability density functions (PDFs), global filters can capture more complex, non-Gaussian PDFs. The point-mass filter (PMF) is a global filter that discretises the state-space into a grid and approximates the PDF as a piecewise constant function, known as the point-mass density (PMD). Unlike particle filters, the PMF produces deterministic estimates: given the same measurements, it will always generate identical outputs.This deterministic nature, combined with its structured representation of the entire distribution,higher robustness and better handling of abrupt changes, makes it valuable in high-reliabilityapplications like navigation systems.Item Testing Platform for Periodic Control(Západočeská univerzita v Plzni, 2025) Myslivec, TomášIndustrial processes face inherent disturbances from material variations, environmentalfactors, and equipment wear. At the same time, modern manufacturing depends on precisioncontrol for applications like robotics, servo systems, and power electronics, demanding robustcontrol solutions that maintain both productivity and quality.The presented Testing Platform for Periodic Control enables safe algorithm testing forsystems with periodic disturbances. Its modular design allows easy reconfiguration for variousdisturbance patterns, while safety features like torque-limited motors and protective enclosuresmake it ideal for education.Item Autoregressive Upscaling of Sparse Single-Cell Data Improves Interpretability(Západočeská univerzita v Plzni, 2025) Honzík, Tomáš; Kuhajda, Lukáš; Georgiev, DanielIntroduction of a novel generative training procedure for modeling single-cell RNA sequencing (scRNA-seq) data based on an autoregressive neural network architecture. Our model sequentially samples UMI-tagged transcripts and effectively captures the complex and sparse distributions inherent in scRNA-seq datasets. This generative framework supports realistic synthetic cell generation, gene expression inpainting, and measurement upscaling. Moreover, the pretrained model serves as a robust foundation for downstre am predictive tasks, such as disease classification. Finally, we propose a novel unsupervised cell-typing approach leveraging the model’s intrinsic generative structure. Cell-type hierarchies naturally emerge by tracing generative sampling paths, offering both interpretability and valuable biological insights.Item Complexity of KIR allele identification pipeline design(Západočeská univerzita v Plzni, 2025) Wolf, Kateřina; Jani, Filip; Georgiev, Daniel; Jindra, Pavel; Holubová, Monika; Houdová, LucieIn Czechia, 1200 people are diagnosed with hemato-oncological disease each year. Treatment of this disease involves hematopoietic stem cell transplantation (HSCT), and its’ success depends on choosing the suitable donor. Although genetic variant in genes other than Human Leukocyte Antigen (HLA) are not primary criteria for the donor selection, their influence can play a role in the patient outcome and it is usually involved in post-transplant conditions (GVL, GVHD). Killer-cell immunoglobulin-like receptor (KIR) genes, affecting HSCT outcome by presence/absence of specific gene, are currently studied for their potential through gene variants (alleles). An analytical pipeline that enables quick and easy assessment of KIR allele is desirable. However, during the pipeline design, it is necessary to address and manage certain issues.Item Semantic Search and Filtering with AI Agents(Springer, 2025) Bulín, Martin; Švec, Jan; Polák, Filip; Šmídl, LubošThe rapid advancement of pre-trained large language models (LLMs) has enabled the creation of innovative applications, especially in natural language processing. This work employs LLMs alongside our in-house technologies to develop an intuitive database search engine that processes natural language queries. The system uses a network of AI agents, including prompted LLMs and single-purpose neural classifiers, to categorize user queries into conditions for filtering individual data sources or direct matches to database entries. Enhanced with a Retrieval-Augmented Generation (RAG) approach, the application allows users to search large databases conversationally through a voice-enabled web-based interface. Currently, in the demo stage, this project shows full pipeline functionality and has been tested with approximately 150 h of transcribed speech data. Initial findings confirm the overall concept of the application.Item An Analytical Design Method for Power System Stabilizers Based on H∞ Specifications(Západočeská univerzita v Plzni, 2025) Brabec, Michal; Dostálek, LukášThis paper presents an analytical design approach for power system stabilizers (PSS) aimed at effective damping of electromechanical oscillations in synchronous generators. The method combines automatic frequency-domain identification based on swept-sine excitation with an analytical controller design method using H∞ specifications. Design constraints, including robustness and damping requirements, are transformed into regions in the parametric plane of fixed-structure controllers, enabling multi-objective and multi-model design without numerical optimization. The approach is applicable to standard PSS structures with lead–lag compensators and washout filters and supports models with uncertainty and varying operating conditions.Item Aspects of density approximation by tensor trains(IEEE, 2025) Ajgl, Jiří; Straka, OndřejPoint-mass filters solve Bayesian recursive relations by approximating probability density functions of a system state over grids of discrete points. The approach suffers from the curse of dimensionality. The exponential increase of the number of the grid points can be mitigated by application of low-rank approximations of multidimensional arrays. Tensor train decompositions represent individual values by the product of matrices. This paper focuses on selected issues that are substantial in state estimation. Namely, the contamination of the density approximations by negative values is discussed first. Functional decompositions of quadratic functions are compared with decompositions of discretised Gaussian densities next. In particular, the connection of correlation with tensor train ranks is explored. Last, the consequences of interpolating the density values from one grid to a new grid are analysed.Item Stone Soup: ADS-B-based Multi-Target Tracking with Stochastic Integration Filter(IEEE, 2025) Hiles, John; Matoušek, Jakub; Blasch, Erik; Niu, Ruixin; Straka, Ondřej; Duník, JindřichThis paper focuses on the multi-target tracking using the Stone Soup framework. In particular, we aim at evaluation of two multi-target tracking scenarios based on the simulated class-B dataset and ADS-B class-A dataset provided by OpenSky Network. The scenarios are evaluated w.r.t. selection of a local state estimator using a range of the Stone Soup metrics. Source code with scenario definitions and Stone Soup set-up are provided along with the paper.Item Interpretable Augmented Physics-Based Model for Estimation and Tracking(IEEE, 2025) Straka, Ondřej; Duník, Jindřich; Closas, Pau; Imbiriba, TalesState-space estimation and tracking rely on accurate dynamical models to perform well. However, obtaining an accurate dynamical model for complex scenarios or adapting to changes in the system poses challenges to the estimation process. Recently, augmented physics-based models (APBMs) appear as an appealing strategy to cope with these challenges where the composition of a small and adaptive neural network with known physics-based models (PBM) is learned on the fly following an augmented state-space estimation approach. A major issue when introducing data-driven components in such a scenario is the danger of compromising the meaning (or interpretability) of estimated states. In this work, we propose a novel constrained estimation strategy that constrains the APBM dynamics close to the PBM. The novel state-space constrained approach leads to more flexible ways to impose constraints than the traditional APBM approach. Our experiments with a radar-tracking scenario demonstrate different aspects of the proposed approach and the trade-offs inherent in the imposed constraints.Item Efficient Gaussian Mixture Filters Based on Transition Density Approximation(IEEE, 2025) Straka, Ondřej; Hanebeck, Uwe D.Gaussian mixture filters for nonlinear systems usually rely on severe approximations when calculating mixtures in the prediction and filtering step. Thus, offline approximations of noise densities by Gaussian mixture densities to reduce the approximation error have been proposed. This results in exponential growth in the number of components, requiring ongoing component reduction, which is computationally complex. In this paper, the key idea is to approximate the true transition density by an axis-aligned Gaussian mixture, where two different approaches are derived. These approximations automatically ensure a constant number of components in the posterior densities without the need for explicit reduction. In addition, they allow a trade-off between estimation quality and computational complexity.Item Asking Questions: an Innovative Way to Interact with Oral History Archives(International Speech Communication Association, 2023) Švec, Jan; Bulín, Martin; Frémund, Adam; Polák, FilipThe paper describes our initial effort to use Transformer-based neural networks for understanding and presenting oral history archives. Such archives of interviews often contain large passages of the interviewee’s speech. Our approach automatically generates relevant questions, which enrich such monotonous parts and allows the listener to better orient in the interview. The generated questions also allow for finding interesting parts of the interview without changing the original meaning of the testimony. We present our working pipeline consisting of a Wav2Vec speech recognizer, BERT-based punctuation detection, T5 asking questions model and BERT-based semantic continuity model.Item The System for Efficient Indexing and Search in the Large Archives of Scanned Historical Documents(Springer, 2023) Bulín, Martin; Švec, Jan; Ircing, PavelThe paper introduces software capable of indexing and searching large archives of scanned historical documents. The system capabilities are demonstrated on the collection containing documents from the archives of the post-Soviet security services. The backend of the system was designed with a focus on flexibility (it is actually already being used for other related tasks) and scalability to larger volumes of data. The graphical user interface design has been consulted with historians interested in using the archived documents and was developed in several iterations, gradually including the changes induced both by the user’s requests and by our improving knowledge about the nature of the processed data.Item Multimodal Low-Cost Robotic Entity based on Raspberry Pi(Západočeská univerzita v Plzni, 2023) Bulín, MartinWith the presence of numerous high-priced robotic entities available in the market, there arises a pressing need to develop low-cost alternatives for proof-of-concept validation, under- scoring the demand for affordable solutions. This study focuses on integrating students’ ma- chine learning projects onto a real robotic platform, facilitating hands-on experience and bridg- ing the theory-practice gap. The primary objective is to develop a versatile robotic device with multiple interfaces as a platform for students’ projects implementation and fundamental ideas verification. By applying their ideas to an embodied entity and testing real-life scenarios, stu- dents can enhance their understanding of complex principles while fostering innovation.Item Design of Efficient Point-Mass Filter with Terrain Aided Navigation Illustration(IEEE, 2023) Matoušek, Jakub; Duník, Jindřich; Brandner, MarekThis paper deals with state estimation of stochastic models with linear state dynamics, continuous or discrete in time. The emphasis is laid on a numerical solution to the state prediction by the time-update step of the grid-point-based point-mass filter (PMF), which is the most computationally demanding part of the PMF algorithm. A novel efficient PMF (ePMF) estimator, unifying continuous and discrete, approaches is proposed, designed, and discussed. By numerical illustrations, it is shown, that the proposed ePMF can lead to a time complexity reduction that exceeds 99.9% without compromising accuracy. The MATLAB® code of the ePMF is released with this paper.
