Parliamentary speech generation presents specific challenges for large language models beyond standard text generation tasks. Unlike general text generation, parliamentary speeches require not only linguistic quality but also political authenticity and ideological consistency. Current language models lack specialized training for parliamentary contexts, and existing evaluation methods focus on standard NLP metrics rather than political authenticity. To address this, we present ParliaBench, a benchmark for parliamentary speech generation. We constructed a dataset of speeches from UK Parliament to enable systematic model training. We introduce an evaluation framework combining computational metrics with LLM-as-a-judge assessments for measuring generation quality across three dimensions: linguistic quality, semantic coherence, and political authenticity. We propose two novel embedding-based metrics, Political Spectrum Alignment and Party Alignment, to quantify ideological positioning. We fine-tuned five large language models (LLMs), generated 28k speeches, and evaluated them using our framework, comparing baseline and fine-tuned models. Results show that fine-tuning produces statistically significant improvements across the majority of metrics and our novel metrics demonstrate strong discriminative power for political dimensions.

This paper presents a novel processing architecture for unmanned aerial vehicles (UAVs), designed to support small search-and-rescue teams operating in post-disaster environments. The proposed system consists of two main components. The primary component is dedicated to vehicle and person detection and tracking. It utilizes a Raspberry Pi 5 paired with a Coral TPU Accelerator to process input from the UAV's main camera, leveraging a YOLOv11n neural network and the KCF tracking algorithm. This capability is essential for following the rescue team's vehicle and locating survivors. When a network connection is available, this subsystem transmits captured images to a remote server for further analysis. The secondary component is a peripheral vision subsystem, powered by a Raspberry Pi Zero 2 W. It processes input from four peripheral cameras connected to a Luxonis OAK-FFC-4P board. It is tasked with identifying critical conditions and communicating the findings to the main processing unit, which in turn alerts the operator. Once appropriate instructions are received, the UAV is redirected to the specified location to perform a focused search for survivors.

Abstract Visual defect recognition and its manufacturing applications have been an upcoming topic in recent AI research. Defect datasets are often severely imbalanced and can be additionally burdened with separating classes of high visual similarity. Although various methods of data augmentation have been proposed to mitigate the class imbalance, they often fail to cope with tinier minority classes or have fidelity issues with smaller defects while, at the same time, needing significant computational resources to train. Also, augmentation based on vector-based oversampling struggles to produce high-fidelity inputs and is hard to apply on custom CNN architectures, which often perform better for this type of problem. Our work presents an image-level oversampling method based on an instance-based image generator that can be applied to any CNN directly during the training process without increasing the order of training time required. It is based on identifying a small number of the most uncertain base samples close to the estimated class boundaries and using them as seeds for augmentation. The resulting images are of high visual quality preserving small class differences, and they also improve the classifier boundary leading to higher recall scores than other state-of-the-art approaches.

The e-Prevention challenge concerns the analysis and processing of long-term continuous recordings of biosignals recorded from wearable sensors, i.e., accelerometers, gyroscopes and heart rate monitors embedded in smartwatches, as well as sleep information and daily step count, in order to extract high-level representations of the wearer’s activity and behavior, termed as digital phenotypes. The ability of these digital phenotypes to quantify behavioral patterns and traits will be evaluated in two different tasks: 1) Person Identification, and 2) Relapse Detection in patients in the psychotic spectrum. The long-term data that will be used in this challenge have been acquired during the course of the e-Prevention project, an innovative integrated system for medical support that facilitates effective monitoring and relapse prevention in patients with mental disorders (i.e, schizophrenia and bipolar disorder). Specifically, the data were continuously collected from patients for a monitoring period of up to 2.5 years, while from the control subgroup for a period of 3 months, constituting one of the largest of its kind ever recorded.

This paper describes a novel web-based platform promoting real-time advanced teleconsultation services on medical imaging. Principles of heterogeneous workflow management systems and state-of-the-art technologies such as the microservices architectural pattern, peer-to-peer networking, and the single-page application concept are combined to build a scalable and extensible platform to aid collaboration among geographically distributed healthcare professionals. The real-time communication capabilities are based on the webRTC protocol to enable direct communication among clients. This paper discusses the conceptual and technical details of the system, emphasizing on its innovative elements.

The abundance of activity trackers and biosignal sensors as well as the evolution of IoT and communication technologies have considerably advanced the concept of Quantified-Self. Nowadays there are several frameworks and applications that realize the concept, focusing though strictly on specific areas, from daily use to professional activities such as sport and healthcare. This work proposes a versatile, cross-domain solution for building quantified-self applications exploiting the capacities for open-design, modularity and extensibility of the AGILE IoT gateway.

In this work, we present a novel web-based platform for real-time teleconsultation services on medical imaging. The introduced platform encompasses the principles of heterogeneous Workflow Management Systems (WFMSs) and the peer-to-peer paradigm to enhance collaboration among healthcare professionals. The novelty of the presented work is that: (i) it combines the concept of client side computing with the microservices architectural pattern to enhance user experience, while increasing the efficiency of healthcare provision, and that (ii) it could be considered the springboard of an ecosystem of Computer-Supported Cooperative Work (CSCW) services in medicine. The paper discusses the implementation details and the technological aspects behind the whole system, along with some initial evaluation results.

BACKGROUND: MicroRNAs are small endogenously expressed non-coding RNA molecules that regulate target gene expression through translation repression or messenger RNA degradation. MicroRNA regulation is performed through pairing of the microRNA to sites in the messenger RNA of protein coding genes. Since experimental identification of miRNA target genes poses difficulties, computational microRNA target prediction is one of the key means in deciphering the role of microRNAs in development and disease. RESULTS: DIANA-microT 3.0 is an algorithm for microRNA target prediction which is based on several parameters calculated individually for each microRNA and combines conserved and non-conserved microRNA recognition elements into a final prediction score, which correlates with protein production fold change. Specifically, for each predicted interaction the program reports a signal to noise ratio and a precision score which can be used as an indication of the false positive rate of the prediction. CONCLUSION: Recently, several computational target prediction programs were benchmarked based on a set of microRNA target genes identified by the pSILAC method. In this assessment DIANA-microT 3.0 was found to achieve the highest precision among the most widely used microRNA target prediction programs reaching approximately 66%. The DIANA-microT 3.0 prediction results are available online in a user friendly web server at http://www.microrna.gr/microT.

The integration of instruments with the grid promotes scientific and business collaboration by allowing shared access to rare and expensive instrumentation, regardless of its physical location. Instruments on the grid are virtualized resources similar to storage and computing devices, and can be part of composite services or workflows concurrently invoked by multiple users in such service-oriented environments. The requirement for real-time, interactive control of instruments poses a need for exclusive or limited access by concurrent clients. Unfortunately, instrument modeling and standardization of instrument reservation and management are still in a very early stage. According to a number of use case requirements, which we review in this work, we propose a standards-compliant model for the representation of instruments on the grid, and we describe a set of primitives for their reservation. The definition of these constructs can be used for the normative description of such reservation capabilities and the future specification of a standardized protocol.

This paper deals with general nested loops and proposes a novel scheduling methodology for reducing the communication cost of parallel programs. General loops contain complex loop bodies (consisting of arbitrary program statements, such as assignments, conditions and repetitions) that exhibit uniform loop-carried dependencies. Therefore it is now possible to achieve efficient parallelization for a vast class of loops, mostly found in DSP, PDEs, signal and video coding. We use computational geometry methods, that exploit efficiently the regularity of nested loops index spaces, in order to significantly reduce the communication cost, which in most cases is the main drawback of parallel programs' performance. Through extensive testing, we show that the proposed method outperforms in all cases the classic cyclic mapping, succeeding to reduce the communication by 15%-35%. This significant reduction of the communication volume makes our method a promising candidate to be incorporated into existing automatic parallel code generation tools

Volunteer computing is a relatively new computing paradigm that is often the only solution to large computational tasks. Due to the dynamic structure of the peer-to-peer systems that usually form the backbone of volunteer computing, obtaining accurate distance metrics using traditional techniques is infeasible. Thus, techniques that could improve system performance, such as dynamic load balancing, are very difficult to implement. We show that the master-worker paradigm used by the overwhelming majority of volunteer computing applications results in certain network topology and protocol characteristics that render distance measurement method with zero network overhead feasible and preferable. Therefore, we present a simple distance measurement method that is based on passive monitoring of application-level traffic. We subsequently test the method using simulation as well as a reference implementation. The paper concludes by summarizing the strong points and limitations of the proposed method and proposing further research directions.

‘Unit execution time’ precedence constrained scheduling (UET) is an NP-complete problem with very few special cases known to be solvable in P-time. In this article we present a practically useful case of UET solvable in P-time: we show that if the task graph is given in levels that are ‘locally’ of in-degree two and of ‘width’ more than 1.55 times the number of processors (plus 1), then an optimal schedule can be found in P-time. Task graphs which represent algebraic computations fall ordinarily in this category. Our algorithm is based on a limited look-ahead technique which allows us to use it in an on-line fashion. In the appendix we give two short NP-completeness proofs which suggest that both ‘width’ and ‘degree’ restrictions are needed to get a polynomially solvable subcase.

This paper compares three similar loop-grouping methods. All methods are based on projecting the n-dimensional iteration space J/sup n/ onto a k-dimensional one, called the projected space, using (n-k) linear independent vectors. The dimension k is selected differently in each method giving various results. The projected space is divided into discrete groups of related iterations, which are assigned to different processors. Two of the methods preserve optimal time completion, by scheduling loop iterations according to the hyperplane method. The theoretical analysis of the experimental results indicates the appropriate method, for specific iteration spaces and target architectures.

This paper presents Chain Grouping, a new low complexity method for the problem of partitioning the loop iteration space into groups with little intercommunication requirements, for mapping onto mesh-connected architectures. First, the iterations are scheduled in time, according to the hyperplane method, taking into consideration the minimum time displacement. Then, the iteration space is divided into discrete groups of related iterations, which are assigned to different processors, while preserving the optimal completion time. Chain Grouping is based on clustering together neighboring uniform chains of iterations, formed by a particular dependence vector. This vector will be proven as the best among all to reduce the total communication requirements. Inside every group, the optimal hyperplane scheduling is preserved and references to intragroup iterations are considerably increased. The partitioned groups are afterward assigned to meshes of processors. The resulting space mapping maximizes processor utilization and cuts down overall communication delays while preserving the optimal hyperplane time schedule.

Attribute grammars have been used extensively in every phase of traditional compiler construction. Recently, it has been shown that they can also be effectively adopted to handle scheduling algorithms in high-level synthesis. Their main advantages are modularity and declarative notation in the development of design automation environments. In this paper, past results are further elaborated and more scheduling techniques are presented and implemented in a flexible environment for the design automation of digital systems. This novel approach can be proven valuable for fast evaluation of new algorithms and techniques in the field.

We propose a fixed size one dimensional VLSI architecture for the parallel parsing of arbitrary context free (CF) grammars, based on Earley's algorithm. The algorithm is transformed into an equivalent double nested loop with loop carried dependencies. We first map the algorithm into a 1D array with unbounded number of cells. The time complexity of this architecture is O(n), which is optimal. We next propose the partitioning into a fixed number of off the shelf processing elements. Two alternative partitioning strategies are presented considering restrictions, not only in the number of the cells, but also in the inner structure of each cell. In the most restricted case, the proposed architecture has time complexity O(n/sup 3//p*k), where p is the number of available cells and the elements inside each cell are at most k.

A sensory uncertainty field (SUF) is a model of the localization uncertainty of a mobile robot. The value of the SUF at a specific robot configuration q expresses the expected uncertainty of the robot at q, as this would be measured by some localization procedure. Path planning over the SUF provides a way for better localization, and thus fewer failures, during navigation. In this paper we extend the original notion of a SUF to unknown and non-stationary environments. We propose a self-organizing neural network model that is capable of building and maintaining an estimation of the SUF while the robot moves around its free space, based on some dynamic localization information, e.g., Kalman filtering. The attractive feature of our algorithm is its capability of handling both unknown and dynamic, i.e., non-stationary, environments. We present a method for polygonal approximation of the resulting SUF by using the Delaunay triangulation.

In order to localize itself a mobile robot tries to match its sensory information at any instant against a prior environment model, the map. A probabilistic map can be regarded as a model that stores at each robot configuration q the probability density function of the sensor readings at q. By combining the knowledge of its current position, the new-coming sensory information, and the probabilistic map the robot is capable of improving its prior position estimate. In this paper we propose a novel sensor model and a method for maintaining a probabilistic map in cases of dynamic environments. When the environment structure changes, the map must adapt to this change by modifying the sensor densities, at the respective configurations. We propose a combined algorithm for map update and robot localization.

Virtual patient records provide a means for integrated access to patient information that may be scattered around different healthcare organizations (or hospital departments). As Intranets provide, among others, secure access to medical information, they constitute an appropriate technological infrastructure for a virtual patient record implementation. In such cases, a security policy can be enforced by combining the security features of the Intranet with the security features of the intraorganizational systems. However, when a workflow system is implemented to automate interorganizational healthcare processes, an additional security layer is needed. An authorization architecture that serves this purpose is presented in this paper.

The paper presents Chain grouping; a new low complexity method for the problem of partitioning the index space into groups with little intercommunication requirements, for mapping onto distributed mesh connected architectures. First the loop iterations are scheduled in time, according to the hyperplane method, taking into consideration the minimum time displacement. Then, the index space is divided into discrete groups of related computations, which are assigned to different processors, while preserving the optimal makespan. The Chain grouping method is based on grouping along a uniform chain of computations, formed by a particular dependence vector. This vector will be proved as the best to reduce the total communication requirements. Inside every group, the optimal hyperplane scheduling is preserved, and the references to intragroup computations are considerably increased. The partitioned groups are afterwards assigned to meshes of processors. The resulting space mapping maximises processor utilisation and cuts down overall communication delays while preserving the optimal hyperplane time schedule.

Since natural language parsing is a computationally intensive task, the parallel parsing of natural language seems a promising choice. This paper describes both the Eu-PAGE (Eurotra PArser GEnerator) meta-compiler for the Eurotra formalism, which is a tool that automatically generates parallel natural language parsers, and the Dialogos parser, which is a parallel parser for the Greek language generated by Eu-PAGE. Parallel parsers generated by Eu-PAGE are based on finite-state machines, employ coarse-grained parallelism and are portably implemented on top of two parallel software platforms: PVM and Orchid. Orchid uses light-weight processes as the basic unit of parallelism, enhanced with advanced operating system facilities. The collected experimental results so far demonstrate satisfactory speed-ups of the parallel implementations compared to the sequential one.

We describe a novel dynamic dramatization method for narrative presentations. This method accepts as input the original story material, along with a description of its plot written in a special-purpose language. It then analyzes the plot to identify interesting dramatic situations in the story. Based on this content analysis, a presentation manager organizes the presentation and enriches it with appropriate multimedia effects. These effects are associated with interesting dramatic situations, and serve to increase suspense and emphasize plot developments in the narrative. Our method can be used for the development of intelligent front-ends to story databases, for directing assistants in computer-based renditions of narrative works, or for real-time direction of interactive entertainment systems. We are integrating this system in an interactive storytelling environment for Greek mythology.

We describe a novel architecture for indoor navigation, based on qualitative representations of the variations in the interactions between the robot and its environment. We use these representations to localize and guide planning and reaction. The system accepts off-line as input a topological diagram of the environment. It then uses numerical simulation to generate a map, describing qualitative variations in the sensor behavior between adjacent regions in space. An off-line planner stores localized navigation information at each point in the map. During execution, an adaptive controller uses a short-term memory to improve its operation. The qualitative nature of our method, along with the localization performed by the topological planner result in a compact map representation and in linear-time performances for position estimation and path planning during execution. This architecture has been tested in simulation. Our results show that the proposed navigation method is tolerant of sensor inaccuracies, both in obstacle detection and orientation.

We describe a novel global path planning method for autonomous qualitative navigation in indoor environments. Global path planning operates on top of a qualitative map of the environment that describes variations in sensor behavior between adjacent regions in space. The method takes into consideration the global topology of the environment and applies a set of criteria that can minimize the errors in the navigational accuracy of a robotic wheelchair. Our approach uses a modified version of the Dijkstra's shortest path algorithm that takes into consideration the curvature of the trajectory and the off-wall distance of the map points. The algorithm computes in real-time a set of optimal paths for reaching the destination. We have tested our global path planning method in simulation in representative indoor environments with above average complexity. Based on these experiments we have determined empirically a set of values for the parameters of the algorithm that almost always lead to the selection of optimal paths in these environments.

The most important issue when parallelizing sequential programs is the efficient assignment of computations into different processing elements. The most extensive, in terms of time execution, part of a program is the nested loops. Too many approaches have been devoted in parallelizing nested loops, and assigning the concurrent partitions of such a loop into different processors. In the past, all methods have been focused upon linear schedules produced by manipulating the reduced dependence graph, which in some cases achieve near optimal solutions. This paper presents a new method of free scheduling loop computations into time, based on task graph scheduling techniques. It will be shown that this schedule is optimal in terms of time, outperforming all linear schedules. Furthermore, in terms of total number of processors, the presented method includes a heuristic refinement of the free schedule which ‘shuffles’ computations into time, without loss of the optimal time performance, to augment the mean processor utilization. In all cases, the proposed method uses less number of processors, while preserving the optimal total execution time. The ‘shuffling’ of computations is based on graph theory approaches, and uses PERT problem techniques. Such scheduling is convenient for parallelizing tools (such as compilers), but has practical interest for shared memory multiprocessor systems, where the communication delay imposed by such non-regular scheduling is of no interest.

A methodology based on functional programming for the systematic synthesis of problem-oriented architectures is presented. The different ways of sequencing the input data are shown to lead to a variety of alternative architectures for the same computational problem (with the same behavioural description). The finally selected architecture satisfies certain user-defined restrictions and maximises given performance criteria. The presented methodology is applicable to a wide range of computing problems and can be used in the process of designing specialised VLSI systems.