Contemporary robotic workplaces are increasingly complex, involving multiple robots, machines, and digital services. Their setup and adaptation demand collaboration across domains, yet domain experts with essential process knowledge are typically non-programmers. Prior research in end-user robot programming has simplified individual task specification, but little is known about how multiple users coordinate and maintain shared awareness when programming together.

Maksim Aparovich, Volha Harytskaya, Vladislav Poritski, Oksana Volchek, Pavel Smrz. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

This paper presents an overview of the Internet of Things (IoT) device design and simulation, with a specific focus on low-power design principles - everything in the context of the LoLiPoP-IoT project. The project aims to enhance IoT device usability by reducing maintenance requirements related to battery recharging or replacement. Another key goal is to significantly decrease the massive waste generated by discarded primary batteries, contributing to more sustainable and user-friendly IoT solutions for the future. The primary focus of this paper is on a custom IoT localization tag, for which we simulate solar cells - ranging from basic modeling to their integration into electrical circuits - and the power consumption of the tag's electronics platform. The analyzed sample platform is built on the nRF52833 microcontroller and the DW3110 ultra-wideband transceiver. We also applied our experimental framework principles to optimize power consumption and extend battery life. Reductions in photovoltaic panel area were achieved for both devices with a 5-year lifespan and fully autonomous tags, though with increased localization latency. Furthermore, this paper demonstrates how IoT devices, including their firmware, can be effectively modeled and simulated using publicly available tools.

The increasing interest in IoT devices poses significant challenges in battery waste, not to mention the effort needed to replace the batteries in remote applications. In the LoLiPoPIoT project, we address these issues by developing energy-efficient IoT platforms that extend the device’s battery life through lowpower design and energy harvesting. However, for a successful design, preliminary measurements and simulations need to be done to demonstrate the feasibility of the utilized technologies and their dimensioning. This paper presents a method of collecting data and creating a novel portable simulation model implemented in a Microsoft Excel spreadsheet. During the data collection phase, we plot detailed energy consumption and income of our device (i.e., asset tracking tag). PC1D simulations of a crystalline-silicon photo-voltaic panel and the device model were utilized to collect data for the final (i.e., composed) model. Our model allows real-time adjustments of key parameters, such as ambient light intensity, battery capacity, and photo-voltaic panel sizing. Our goal is to deliver a solution that allows our partner to conduct basic experiments on their own computers with a user-friendly interface for parameter selection and result visualization. Our results indicate that for one application area, a $25 \mathrm{~cm}^{2}$ photovoltaic panel is needed, while for the other, a $23 \mathrm{~cm}^{2}$ panel is enough.

Abstract We present BenCzechMark (BCM), the first comprehensive Czech language benchmark designed for large language models, offering diverse tasks, multiple task formats, and multiple evaluation metrics. Its duel scoring system is grounded in statistical significance theory and uses aggregation across tasks inspired by social preference theory. Our benchmark encompasses 50 challenging tasks, with corresponding test datasets, primarily in native Czech, with 14 newly collected ones. These tasks span 8 categories and cover diverse domains, including historical Czech news, essays from pupils or language learners, and spoken word. Furthermore, we collect and clean BUT-Large Czech Collection, the largest publicly available clean Czech language corpus, and use it for (i) contamination analysis and (ii) continuous pretraining of the first Czech-centric 7B language model with Czech-specific tokenization. We use our model as a baseline for comparison with publicly available multilingual models. Lastly, we release and maintain a leaderboard with existing 50 model submissions, where new model submissions can be made at https://huggingface.co/spaces/CZLC/BenCzechMark.

The LoLiPoP-IoT project aims to pioneer Long Life Power Platforms for IoT to extend battery life, minimize maintenance, and facilitate installation within existing environments. With a focus on supporting an inclusive ecosystem of developers, integrators, coordinators, and users, the project's Grand Objectives encompass a range of aims, including providing long-lasting battery solutions, reducing battery waste, enhancing asset tracking and predictive maintenance, and improving energy efficiency in buildings. These objectives are realized through nine selected practical applications across three primary domains: Asset Tracking, Condition Monitoring and Predictive Maintenance, and Energy Efficiency and Comfort in Buildings. Expected impacts of the LoLiPoP-IoT project include significantly extended battery life, reduced maintenance overhead, decreased costs associated with asset location, improved asset management efficiency, enhanced building comfort with reduced energy consumption, and substantial revenue generation for industry partners. The project's strategic objectives are notably harmonized with key EU initiatives outlined in the Green Deal, Circular Economy, and the New Industrial Strategy for Europe.

The paper presents an approach for handheld augmented reality in constrained industrial environments, where it might be hard or even impossible to reach certain poses within a workspace. Therefore, a user might be unable to see or interact with some digital content in applications like visual robot programming, robotic program visualizations, or workspace annotation. To overcome this limitation, we propose a temporal switching to a non-immersive virtual reality that allows the user to see the virtual counterpart of the workspace from any angle and distance, where the viewpoint is controlled using a unique combination of on-screen controls complemented by the physical motion of the handheld device. Using such a combination, the user can position the virtual camera roughly to the desired pose using the on-screen controls and then continue working just as in augmented reality. To explore how people would use it and what the benefits would be over pure augmented reality, we chose a representative task of object alignment and conducted a study. The results revealed that mainly physical demands, which is often a limiting factor for handheld augmented reality, could be reduced and that the usability and utility of the approach are rated as high. In addition, suggestions for improving the user interface were proposed and discussed.

The paper presents an approach for handheld augmented reality in constrained industrial environments, where it might be hard or even impossible to reach certain poses within a workspace. Therefore, a user might not be able to see or interact with some digital content in applications like visual robot programming, robotic program visualizations, or workspace annotation. To overcome this limitation, we propose a temporal switching to a non-immersive virtual reality, enabling the user to see the workspace from any angle and distance. To explore how people would use it and what the benefits would be over pure augmented reality, we chose a representative task of object alignment and conducted a study. The results revealed that mainly physical demands, which is often a limiting factor for handheld augmented reality, could be reduced and that the usability and utility of the approach are rated as high. In the next iteration, we want to investigate other possibilities of controlling the viewpoint in the virtual environment, as the current approach has potential for improvements.

The problem of selecting virtual objects within augmented reality on handheld devices has been tackled multiple times. However, evaluations were carried out on purely synthetic tasks with uniformly placed homogeneous objects, often located on a plane and with none or low occlusions. This paper presents two novel approaches to indirect object selection dealing with highly occluded objects with large spatial distribution variability and heterogeneous size and appearance. The methods are de-signed to enable long-term usage with a tablet-like device. One method is based on a spatially anchored hierarchy menu, and the other utilizes a crosshair and a side menu that shows candidate objects according to a custom-developed metric. The proposed approaches are compared with direct touch in the context of spatial visual programming of collaborative robots problem, on a realistic workplace and a common robotic task. The preliminary evaluation indicates that the main benefit of the proposed indirect methods could be their higher precision and higher selection confidence for the user.

The OCCAM project (Optical Character recognition, ClassificAtion & Machine Translation) aims at integrating the CEF (Connecting Europe Facility) Automated Translation service with image classification, Translation Memories (TMs), Optical Character Recognition (OCR), and Machine Translation (MT). It will support the automated translation of scanned business documents (a document format that, currently, cannot be processed by the CEF eTranslation service) and will also lead to a tool useful for the Digital Humanities domain.

In this paper, we present the MegaM@Rt2 ECSEL project and discuss in details our approach for fostering collaboration in this project. We choose to use an internal hackathon approach that focuses on technical collaboration between case study owners and tool/method providers. The novelty of the approach is that we organize the technical workshop at our regular project progress meetings as a challenge-based contest involving all partners in the project. Case study partners submit their challenges related to the project goals and their use cases in advance. These challenges are concise enough to be experimented within approximately 4 hours. Teams are then formed to address those challenges. The teams include tool/method providers, case study owners and researchers/developers from other consortium members. On the hackathon day, partners work together to come with results addressing the challenges that are both interesting to encourage collaboration and convincing to continue further deeper investigations. Obtained results demonstrate that the hackathon approach stimulated knowledge exchanges among project partners and triggered new collaborations, notably between tool providers and use case owners.

This paper proposes an intuitive approach for collaborative robot end-user programming using a combination of interactive spatial augmented reality (ISAR) and headmounted display (HMD). It aims to reduce user's workload and to let the user program the robot faster than in classical approaches (e.g. kinesthetic teaching). The proposed approach, where user is using a mixed-reality HMD - Microsoft HoloLens - and touch-enabled table with SAR projected interface as input devices, is compared to a baseline approach, where robot's arms and a touch-enabled table are used as input devices. Main advantages of the proposed approach are the possibility to program the collaborative workspace without the presence of the robot, its speed in comparison to the kinesthetic teaching and an ability to quickly visualize learned program instructions, in form of virtual objects, to enhance the users' orientation within those programs. The approach was evaluated on a set of 20 users using the within-subject experiment design. Evaluation consisted of two pick and place tasks, where users had to start from the scratch as well as to update the existing program. Based on the experiment results, the proposed approach is better in qualitative measures by 33.84% and by 28.46% in quantitative measures over the baseline approach for both tasks.

This paper presents a novel approach to interaction between human workers and industrial collaborative robots. The proposed approach addresses problems introduced by existing solutions for robot programming. It aims to reduce the mental demands and attention switches by centering all interaction in a shared workspace, combining various modalities and enabling interaction with the system without any external devices. The concept allows simple programming in the form of setting program parameters using spatial augmented reality for visualization and a touch-enabled table and robotic arms as input devices. We evaluated the concept utilizing a user experience study with six participants (shop-floor workers). All participants were able to program the robot and to collaborate with it using the program they parametrized. The final goal is to create a distraction-free, usable and low-effort interface for effective human-robot collaboration, enabling any ordinary skilled worker to customize the robot's program to changes in production or to personal (e.g. ergonomic) needs.

Recently, there is an increasing tendency to embed functionalities for recognizing emotions from user-generated media content in automated systems such as call-centre operations, recommendations, and assistive technologies, providing richer and more informative user and content profiles. However, to date, adding these functionalities was a tedious, costly, and time-consuming effort, requiring identification and integration of diverse tools with diverse interfaces as required by the use case at hand. The MixedEmotions Toolbox leverages the need for such functionalities by providing tools for text, audio, video, and linked data processing within an easily integrable plug-and-play platform. These functionalities include: 1) for text processing: emotion and sentiment recognition; 2) for audio processing: emotion, age, and gender recognition; 3) for video processing: face detection and tracking, emotion recognition, facial landmark localization, head pose estimation, face alignment, and body pose estimation; and 4) for linked data: knowledge graph integration. Moreover, the MixedEmotions Toolbox is open-source and free. In this paper, we present this toolbox in the context of the existing landscape, and provide a range of detailed benchmarks on standard test-beds showing its state-of-the-art performance. Furthermore, three real-world use cases show its effectiveness, namely, emotion-driven smart TV, call center monitoring, and brand reputation analysis.

A major challenge for the European electronic industry is to enhance productivity while reducing costs and ensuring quality in development, integration and maintenance. Model-Driven Engineering (MDE) principles and techniques have already shown promising capabilities but still need to scale to support real-world scenarios implied by the full deployment and use of complex electronic components and systems. Moreover, maintaining efficient traceability, integration and communication between two fundamental system life-time phases (design time and runtime) is another challenge facing scalability of MDE. This paper presents an overview of the ECSEL project entitled "MegaModelling at runtime -- Scalable model-based framework for continuous development and runtime validation of complex systems" (MegaM@Rt2), whose aim is to address the above mentioned challenges facing MDE. Driven by both large and small industrial enterprises, with the support of research partners and technology providers, MegaM@Rt2 aims to deliver a framework of tools and methods for: 1) system engineering/design & continuous development, 2) related runtime analysis and 3) global model & traceability management, respectively. The diverse industrial use cases (covering domains such as aeronautics, railway, construction and telecommunications) will integrate and apply such a framework that shall demonstrate the validation of the MegaM@Rt2 solution.

This paper examines user interface options and interaction patterns evinced in tools for computer-assisted semantic enrichment of text. It focuses on advanced annotation tasks such as hierarchical annotation of complex relations and linking entities with highly ambiguous names and explores how decisions on particular aspects of annotation interfaces influence the speed and the quality of computer-assisted human annotation processes. Reported experiments compare the 4A annotation system, designed and implemented by our team, to RDFaCE and GATE tools that all provide advanced annotation functionality. Results show that users are able to reach better consistency of event annotations in less time when using the 4A editor. A set of experiments is then conducted that employ 4A’s high flexibility and customizability to find an optimal amount of displayed information and its presentation form to reach best results in linking entities with highly ambiguous names. The last set of experiments then proves that 4A’s particular way of implementing the concept of semantic filtering speeds up event annotation processes and brings higher consistency when compared to alternative approaches.

This paper presents results of an exploratory study comparing various modalities employed in an industrial-like robot-human shared workplace. Experiments involved 39 participants who used a touch table, a touch display, hand gestures, a 6D pointing device, and a robot arm to show the robot how to assemble a simple product. To rule out a potential dependence of results on the number of misrecognized actions (resulting, e.g., from unreliable gesture recognition), a controlled amount of interaction errors was introduced. A Wizard-of-Oz setting with three user groups differing in the amount of simulated recognition errors helped us to show that hand gestures and 6D pointing are the fastest modalities that are also generally preferred by users for setting parameters of certain robot operations.

A typical high-end film production generates several terabytes of data per day, either as footage from multiple cameras or as background information regarding the set (laser scans, spherical captures, etc). This paper presents solutions to improve the integration of the multiple data sources, and understand their quality and content, which are useful both to support creative decisions on-set (or near it) and enhance the postproduction process. The main cinema specific contributions, tested on a multisource production dataset made publicly available for research purposes, are the monitoring and quality assurance of multicamera set-ups, multisource registration and acceleration of 3-D reconstruction, anthropocentric visual analysis techniques for semantic content annotation, and integrated 2-D–3-D web visualization tools. We discuss as well improvements carried out in basic techniques for acceleration, clustering and visualization, which were necessary to deal with the very large multisource data, and can be applied to other big data problems in diverse application fields.

In the modern digital cinema production, extremely large volumes (in order of 10s of TB) of footage data are captured every day. The process of cataloging and reviewing such footage is nowadays largely manual and time consuming process. In our work, we aim at technical quality aspects, such as correct exposure, color compatibility of adjacent shots, and focusing. The main goal is to assist the reviewing process by providing shot quality meta-data and possibly ordering or even culling significant portion of the data from the review, as better quality shot of the same scene exists. However, in order to meaningfully compare technical quality, temporal shot synchronization needs to be performed first. We propose a fast and robust method for time synchronization of video sequences, capturing similar scenes, which arise naturally in digital cinema production. The method is tested on an extensive library of sequences and its performance is evaluated. We further present a preview of application of the proposed method to detecting focusing errors.

Many estimation problems in robotics rely on efficiently solving nonlinear least squares (NLS). For example, it is well known that the simultaneous localisation and mapping (SLAM) problem can be formulated as a maximum likelihood estimation (MLE) and solved using NLS, yielding a mean state vector. However, for many applications recovering only the mean vector is not enough. Data association, active decisions, next best view, are only few of the applications that require fast state covariance recovery. The problem is not simple since, in general, the covariance is obtained by inverting the system matrix and the result is dense. The main contribution of this paper is a novel algorithm for fast incremental covariance update, complemented by a highly efficient implementation of the covariance recovery. This combination yields to two orders of magnitude reduction in computation time, compared to the other state of the art solutions. The proposed algorithm is applicable to any NLS solver implementation, and does not depend on incremental strategies described in our previous papers, which are not a subject of this paper.

Stream processing is a paradigm evolving in response to well-known limitations of widely adopted MapReduce paradigm for big data processing, a hot topic of today's computer world. Moreover, in the field of computation facilities, heterogeneity of data processing clusters, intended or unintended, is starting to be relatively common. This paper deals with scheduling problems and decisions in stream processing on heterogeneous clusters. It brings an overview of current state of the art of stream processing on heterogeneous clusters with focus on resource allocation and scheduling. Basic scheduling decisions are discussed and demonstrated on naive scheduling of a sample application. The paper presents a proposal of a novel scheduler for stream processing frameworks on heterogeneous clusters, which employs design-time knowledge as well as benchmarking techniques to achieve optimal resource-aware deployment of applications over the clusters and eventually better overall utilization of the cluster.

A large number of robotic, computer vision and computer graphics applications rely on efficiently solving the associated sparse linear systems. Simultaneous localization and mapping (SLAM), structure from motion (SfM), non-rigid shape recovery, and elastodynamic simulations are only few examples in this direction. In general, these problems are nonlinear and the solution can be approximated by incrementally solving a series of linearized problems. In some applications, the size of the system considerably affects the performance, especially when the sparsity is low. This paper exploits the block structure of such problems and offers very efficient solutions to manipulate block matrices within iterative nonlinear solvers. The resulting method considerably speeds-up the execution of the implementation of the nonlinear optimization problem. In this work, in particular, we focus our effort on testing the method on SLAM applications, but the applicability of the technique remains general. Our implementation outperforms the state of the art SLAM implementations on all tested datasets. In incremental mode, where a larger portion of time is spent in updating the system, our implementation is on average two times faster than the others.

This paper deals with a system for extracting information from scientific papers. We analyze drawbacks of an existing implementation running on the N1 Grid Engine. Reasons for moving extraction to the Cloud are presented next. The architecture of the Cloud port is discussed and the links to the API and the platform developed within the mOSCAIC project are elaborated.

RDF Schema is a basic and yet very important language for specifying ontologies in the context of Semantic Web. Ontologies can be used to obtain more information from that explicitly stated. Traditionally, the process of revealing implicit knowledge, known as inference or reasoning, is realised by a reasoner -- a component which either processes data to infer all conclusions in advance or, given a query, infers all implicit answers during a query evaluation process. In this paper, we present an alternative approach based on structuring data in a database according to underlying ontologies. This knowledge structure is employed for directing query evaluation to data relevant for the given query. Thus, this method reduces reasoning to retrieving. We show that the proposed schema preserves the semantics of RDF and RDFS in standard use cases.

This paper deals with a scene pre-processing task - depth image segmentation. Efficiency and accuracy of several methods for depth map segmentation are explored. To meet real-time capable constraints, state-of-the-art techniques needed to be modified. Along with these modifications, new segmentation approaches are presented which aim at optimizing performance characteristics. They benefit from an assumption of human-made indoor environments by focusing on detection of planar regions. All methods were evaluated on datasets with manually annotated real environments. A comparison with alternative solutions is also presented.

SRS is a European research project for building robust personal assistant robots using ROS (Robotic Operating System) and Care-O-bot (COB) 3 as the initial demonstration platform. In this paper, experience gained while building the SRS system is presented. A main contribution of the paper is the SRS autonomous control framework. The framework is divided into two parts. First, it has an automatic task planner, which initialises actions on the symbolic level. The planner produces proactive robotic behaviours based on updated semantic knowledge. Second, it has an action executive for coordination actions at the level of sensing and actuation. The executive produces reactive behaviours in well-defined domains. The two parts are integrated by fuzzy logic based symbolic grounding. As a whole, they represent the framework for autonomous control. Based on the framework, several new components and user interfaces are integrated on top of COB's existing capabilities to enable robust fetch and carry in unstructured environments. The implementation strategy and results are discussed at the end of the paper.