Description Language Research Articles

Assessing and Analyzing the Correctness of GitHub Copilot’s Code Suggestions

AI programming has become a popular topic in recent years. Code suggestion, with code suggestion being a key capability of AI programming. Copilot, an “AI programmer” that provides code suggestions from natural language descriptions, has been launched by GitHub and OpenAI. By far, Copilot has been widely used by millions of developers. However, little work has systematically evaluated the correctness of Copilot's suggestions. We conducted an empirical study on all 2,033 LeetCode problems to assess Copilot's code generation across four mainstream languages: C, Java, JavaScript, and Python. We have found that: 1) 70.0% of problems received at least one correct suggestion, with language-specific rates of 29.7% (C), 57.7% (Java), 54.1% (JavaScript), and 41.0% (Python); 2) Correctness decreases as problem difficulty increases, with acceptance rates of 89.3% (Easy), 72.1% (Medium), and 43.4% (Hard); 3) Acceptance rates vary across problem domains from 49.5% to 90.1%, while Graph problems challenge C and Python most, and Prefix Sum and Heap challenge Java and JavaScript most; 4) For the incorrect suggestions, we further summarize 17 types of error reasons accounting for their incorrectness and analyzed possible causes for why these errors occur. We believe our study can provide valuable insights into Copilot's capabilities and limitations.

An Approach Integrating Model-Based Systems Engineering, IoT, and Digital Twin for the Design of Electric Unmanned Autonomous Vehicles

This article proposes a novel methodology aimed at streamlining the system’s development process. By examining existing state-of-the-art approaches and the capabilities inherent in Model-Based Systems Engineering (MBSE) tools, the article introduces a methodology centered around transforming a descriptive Systems Modeling Language (SysML) model into a digital twin. This virtual representation of the physical asset leverages real-time data and simulations to mirror its behavior and characteristics. When integrated with MBSE, this synergy allows for a comprehensive and dynamic approach, enhancing innovation by providing a holistic and adaptable framework for designing, analyzing, and optimizing complex systems throughout their lifecycle. The practical application of this Real-Time Communication and Data Acquisition (RT-CDA) methodology is implemented in a context and operational scenario of an electric unmanned autonomous vehicle employing both Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HITL) approaches. The methodology empowers systems engineers to iteratively update and refine their system model’s fidelity based on real-world testing insights. The article specifically demonstrates the real-time communication capabilities achieved between an electric unmanned autonomous vehicle (a physical asset) and a descriptive (SysML) model, illustrating the real-time data aspect integral to the concept of a digital twin. This study serves as a foundation for future endeavors, envisioning real-time communication among virtual and physical models to construct comprehensive digital twins of complex systems to predict behavior and performance.

Automatically generated place descriptions for accurate location identification: a hybrid approach with rule-based methods and LLM

ABSTRACT This paper explores the potential of using machine-generated descriptions to characterize a place in a way that humans can identify it. It presents a hybrid approach for generating place descriptions by combining rule-based generation of spatial relation facts with a LLM that converts these facts into natural language descriptions. The study focuses on urban areas and street-level scale, using OpenStreetMap as the primary data source. The rule-based method is informed by an experimental study that analyzed human-made place descriptions to understand reference object types used, their quantities, distances, and spatial relations. An experiment is carried out to assess the quality of machine-generated descriptions compared to human-made descriptions in a place identification task. The evaluation involved 70 participants identifying locations based on both human and machine-generated descriptions across a 200-hectare urban area. The results show that the same average identification accuracy was not achieved. However, the proposed method reached lower variance and the difference in accuracy is not substantial enough to impede place identification in the anticipated use cases. The method shows promise for applications in navigation systems, virtual assistants, and location-based services, particularly in situations where visual media cannot be used.

Easy and Straightforward FPGA Implementation of Model Predictive Control Using HDL Coder

Model Predictive Control (MPC) is widely adopted for power electronics converters due to its ability to optimize system performance under dynamic constraints. However, its FPGA implementation remains challenging due to the complexity of Hardware Description Language (HDL) programming. This paper addresses this challenge by introducing a straightforward methodology that simplifies FPGA implementation using MATLAB Simulink HDL Coder. It is shown that HDL Coder yields favorable synthesis outcomes, both in terms of area and time, compared to hand-coded HDL. Notably, the proposed method achieves a significantly reduced sampling step for the MPC algorithm—down to 32 ns—marking a substantial improvement over state-of-the-art implementations. The Integrated Logic Analyzer (ILA) IP available in the Vivado tool is used during the HIL testing phase to facilitate the real-time observation and analysis required for debugging and confirming the FPGA-implemented controller performance. Additionally, this paper discusses the advantages of utilizing HDL Coder for simplifying the FPGA programming process in power electronics applications and addresses the design challenges encountered using this methodology.

Supporting vision-language model few-shot inference with confounder-pruned knowledge prompt.

Vision-language models are pre-trained by aligning image-text pairs in a common space to deal with open-set visual concepts. Recent works adopt fixed or learnable prompts, i.e., classification weights are synthesized from natural language descriptions of task-relevant categories, to reduce the gap between tasks during the pre-training and inference phases. However, how and what prompts can improve inference performance remains unclear. In this paper, we explicitly clarify the importance of incorporating semantic information into prompts, while existing prompting methods generate prompts without sufficiently exploring the semantic information of textual labels. Manually constructing prompts with rich semantics requires domain expertise and is extremely time-consuming. To cope with this issue, we propose a knowledge-aware prompt learning method, namely Confounder-pruned Knowledge Prompt (CPKP), which retrieves an ontology knowledge graph by treating the textual label as a query to extract task-relevant semantic information. CPKP further introduces a double-tier confounder-pruning procedure to refine the derived semantic information. Adhering to the individual causal effect principle, the graph-tier confounders are gradually identified and phased out. The feature-tier confounders are eliminated by following the maximum entropy principle in information theory. Empirically, the evaluations demonstrate the effectiveness of CPKP in few-shot inference, e.g., with only two shots, CPKP outperforms the manual-prompt method by 4.64% and the learnable-prompt method by 1.09% on average.

From Heritage Building Information Modelling Towards an ‘Echo-Based’ Heritage Digital Twin

Since the late 2000s, numerous studies have focused on the application of Heritage Building Information Modelling (HBIM) processes and technologies for the documentation of the historic built environment. Many of these studies have focused on the use of BIM software tools to generate intelligent 3D models using information gathered from a range of data capture techniques including laser scanning and photogrammetry. While this approach effectively preserves existing or partially extant heritage, it faces limitations in reconstructing lost or poorly documented structures. The aim of this study is to develop a novel approach to complement the existing tangible-based HBIM methods, towards an ‘Echo-based’ Heritage Digital Twin (EH-DT) an early-stage digital representation that leverages intangible, memory-based oral descriptions (or echoes) and AI text-to-image generation techniques. The overall methodology for the research presented in this paper proposes a three-phase framework. Phase 1: engineering a standardised heritage prompt template, Phase 2: creation of the Architectural Heritage Transformer, and Phase 3: implementing an AI text-to-image generation toolkit. Within these phases, intangible data, including collective memories (or oral histories) of people who had first-hand experience with the building, provide ‘echoes’ of past form. These can then be converted using a novel ‘Architectural Heritage Transformer’ (AHT), which converts plain language descriptions into architectural terminology through a generated taxonomy. The output of the AHT forms input for a pre-created standardised heritage prompt template for use in AI diffusion models. While the current EH-DT framework focuses on producing 2D visual representations, it lays the foundation for potential future integration with HBIM models or digital twin systems. However, the reliance on generative AI introduces potential risks of inaccuracies due to speculative outputs, necessitating rigorous validation and iterative refinement to ensure historical and architectural credibility. The findings indicate the potential of AI to extend the current HBIM paradigm by generating images of ‘lost’ heritage buildings, which can then be used to enhance and augment the more ‘traditional’ HBIM process.

A clip-based dual-stream method for text based vehicle search

A text-based vehicle search refers to a system where users can find vehicles or route information by entering text-based queries. The primary objective of text-based vehicle search is to identify the most relevant vehicle in a given dataset using a natural language description as a query. This approach leverages natural language processing (NLP) to understand and interpret description queries and provide relevant results. Despite significant progress, this task still faces several challenges due to the complexity and diversity of natural language, as well as inherent difficulties in the vision domain. Moreover, few studies have focused on tracked-vehicle retrieval, where vehicle tracklets are considered instead of single images. In this paper, we propose a novel framework for natural language-based tracked-vehicle retrieval based on CLIP model, one of the most effective models for image-text matching task. This framework leverages both appearance and motion information to enhance the matching accuracy of vehicle tracklet retrieval. Some experiments are conducted on the CityFlow-NL dataset, provided by the 6-th AI City Challenge, an annual competition. The results are comparable to state-of-the-art methods, achieving an MRR score of 46.63%, Rank@5 of 67.02%, and Rank@10 of 81.82%

The competitive esports physiological, affective, and video dataset

Esports refers to competitive video gaming where individuals compete against each other in organized tournaments for prize money. Here, we present the Competitive Esports Physiological, Affective, and Video (CEPAV) dataset, in which 300 male Counter Strike: Global Offensive gamers participated in a study aimed at optimizing affect during esports tournament1. The CEPAV dataset includes (1) physiological data, capturing the player’s cardiovascular responses from before, during, and after over 3000 CS: GO matches; (2) self-reported affective data, detailing the affective states experienced before gameplay; and (3) video data, providing a visual record of 552 in-laboratory gaming sessions. We also collected (affect-related) individual differences measures (e.g., well-being, ill-being) across six weeks in three waves. The self-reported affective data also includes gamers’ natural language descriptions of gaming affective situations. The CEPAV dataset provides a comprehensive resource for researchers and analysts seeking to understand the complex interplay of physiological, affective, and behavioral factors in esports and other performance contexts.

TinyHLS: a novel open source high level synthesis tool targeting hardware accelerators for artificial neural network inference.

In recent years, wearable devices such as smartwatches and smart patches have revolutionized biosignal acquisition and analysis, particularly for monitoring electrocardiography (ECG). However, the limited power supply of these devices often precludes real-time data analysis on the patch itself. This paper introduces a novel Python package, tinyHLS (High Level Synthesis), designed
to address these challenges by converting Python-based AI models into platform-independent hardware description language (HDL) code accelerators. Specifically designed for convolutional neural networks (CNNs), tinyHLS seamlessly integrates into the AI developer's workflow in Python TensorFlow Keras. Our methodology leverages a template-based hardware compiler that ensures flexibility, efficiency, and ease of use. In this work, tinyHLS is first-published featuring templates for several layers of neural networks, such as dense, convolution, max and global average pooling. In the first version, rectified linear unit (ReLU) is supported as activation. It targets one-dimensional data, with a particular focus on time series data. The generated accelerators are validated in detecting atrial fibrillation (AF) on electrocardiogram (ECG) data, demonstrating significant improvements in processing speed (62-fold) and energy efficiency (4.5-fold). Quality of code and synthesizability are ensured by validating the outputs with commercial ASIC design tools. Importantly, tinyHLS is open-source and does not rely on commercial tools, making it a versatile solution for both academic and commercial applications. The paper also discusses the integration with an open-source RISCV and potential for future enhancements of tinyHLS, including its application in edge servers and cloud computing. The source code is available on GitHub: https://github.com/Fraunhofer-IMS/tinyHLS.

An effective video captioning based on language description using a novel Graylag Deep Kookaburra Reinforcement Learning

Video captioning exhibits a complex challenge, particularly due to the increased subject intensity within videos compared to image caption generation. The presence of redundant visual information in video data adds complexity for captioners, making it difficult to simplify various content and eliminate irrelevant elements. Additionally, this redundancy often results in misalignment with equivalent visual semantics in the ground truth, further complicating the video captioning process. In response to these challenges, this research introduces the Graylag Deep Kookaburra Reinforcement Learning (GDKRL) framework, which enhances video captioning through a multi-stage process. First, object detection is performed using the single-shot multi-box detector with generalized intersection over union for accurate object tracking and similarity calculation. Next, the gazelle autoencoder extracts and fuses features from video frames, integrating complex visual and temporal information into a unified representation. The residual convolved dual sparse graph attention network then generates detailed and contextually rich language descriptions by applying dual sparse attention mechanisms and residual convolutions. Finally, hybrid graylag and kookaburra optimization refine the captioning process, producing comprehensive and precise textual descriptions of the video content. Extensive experiments on MSVD yielded 81.79 BLEU-4, 51.2 METEOR, 133.3 CIDEr and 81.7 ROUGE-L; VATEX achieved 62.29 BLEU-4, 51.2 METEOR, 110.2 CIDEr and 78.45 ROUGE-L; and the MSR-VTT dataset obtained 44.52 BLEU-4, 33.35 METEOR, 63.9 CIDEr and 68.9 ROUGE-L, demonstrating that the proposed technique significantly outperforms previous approaches and highlights its effectiveness.

MKER: multi-modal knowledge extraction and reasoning for future event prediction

Humans can predict what will happen shortly, which is essential for survival, but machines cannot. To equip machines with the ability, we introduce the innovative multi-modal knowledge extraction and reasoning (MKER) framework. This framework combines external commonsense knowledge, internal visual relation knowledge, and basic information to make inference. This framework is built on an encoder-decoder structure with three essential components: a visual language reasoning module, an adaptive cross-modality feature fusion module, and a future event description generation module. The visual language reasoning module extracts the object relationships among the most informative objects and the dynamic evolution of the relationship, which comes from the sequence scene graphs and commonsense graphs. The long short-term memory model is employed to explore changes in the object relationships at different times to form a dynamic object relationship. Furthermore, the adaptive cross-modality feature fusion module aligns video and language information by using object relationship knowledge as guidance to learn vision-language representation. Finally, the future event description generation module decodes the fused information and generates the language description of the next event. Experimental results demonstrate that MKER outperforms existing methods. Ablation studies further illustrate the effectiveness of the designed module. This work advances the field by providing a way to predict future events, enhance machine understanding, and interact with dynamic environments.

A framework for reviewing the results of automated conversion of structured organic synthesis procedures from the literature

We propose a framework to edit automatically converted chemical description language (χDL) from organic synthesis procedure texts, supporting human reviewers by providing annotated text.

Custom TTA Operations for Accelerating the Ascon Encryption Algorithm

Lightweight cryptography is becoming increasingly important in modern applications, especially in resource-constrained environments such as Internet of Things (IoT) devices, embedded systems and mobile platforms. The Ascon encryption algorithm is a modern, secure and efficient cryptographic scheme that meets the demands of low-power devices. However, some steps of the algorithm are computationally intensive, leading to performance issues. In this study, custom operations are proposed to accelerate the Ascon encryption algorithm on Transport-Triggered Architecture (TTA) processors. In order to make more efficient use of hardware resources, the custom operations are designed to have low complexity and high efficiency. The OpenASIP tool was employed to integrate the operations into a general purpose 64-bit TTA processor. The resulting application-specific core was implemented in Hardware Description Language (HDL) and synthesised for FPGA. The performance gain is analysed for different transport bus configurations. The results obtained show that the Ascon-AEAD128 encryption and decryption phases are accelerated by 38% to 50%. When evaluated together with the synthesis results, a significant performance gain was achieved with a very reasonable increase in hardware resources. The study also emphasises that the TTA is a suitable method for accelerating cryptographic applications that require low power consumption and high efficiency.

Reception of Danielius Kleinas’s Ideas in Povilas Frydrichas Ruigys’s Grammar Anfangsgründe einer Littauischen Grammatick (1747)

The article analyses the reception of Danielius Kleinas’s ideas in Povilas Frydrichas Ruigys’s grammar Anfangsgründe einer Littauischen Grammatick (1747, hereinafter RG). Two aims were pursued: (1) to single out Kleinas’s ideas that Ruigys found relevant and suitable for describing the language of the mid-eighteenth century and (2) to identify the intentions of Ruigys in adopting them. Using an analytical interpretive approach based on philological analysis, it is shown that Kleinas’s Grammatica Litvanica (1653, hereinafter KG) was a particularly important source. The need of the Age of Enlightenment to learn about the world and to educate society is highlighted in discussing the motives for adopting and adapting ideas. The study of the reception of Kleinas’s ideas leads to the following conclusions: 1. Frequent references to KG, both explicit and implicit, point to a similar school of linguistic thought and the acceptability of its ideas. 2. KG was a particularly important source in two respects: (1) selecting a dialect for grammar and distinguishing it from others and (2) developing an extensive section on syntax. 3. By adopting Kleinas’s key idea of choosing one dialect as the basis for the description of the Lithuanian language, Ruigys expanded its area to two counties: Įsrutis (Ger. Insterburg) and Ragainė (Ger. Ragnit). He considered the other dialects in Prussia and Lithuania Major as secondary. Ruigys was the first to classify Lithuanian dialects by the opposition main : secondary or centre : periphery. 4. By placing emphasis on the stability, universality, and prestige of the main dialect in relation to other dialects, Ruigys continued the tradition of forming standard language on the basis of one dialect, which was initiated by Kleinas. In the eighteenth century, this was relevant for the written form of the language. 5. By introducing secondary dialects, he aimed to facilitate the integration of individuals appointed to priesthood in a new linguistic environment. This was important for the spoken form of the language. 6. In preparing a grammar for learning Lithuanian by non-Lithuanians, Ruigys paid considerable attention to syntax. By adopting Kleinas’s ideas and expanding them, he sought to develop the ability to produce a coherent text. The ability to generate well-formed sentences is important for both written and spoken language. 7. One of the characteristics of the language teaching methodology developed by Ruigys was the development of linguistic communicative competence. In the Age of Enlightenment, this was particularly necessary for the organisation of public education through churches and schools. Taking up earlier ideas and adapting them to the needs of the eighteenth century, Ruigys developed a tool for non-Lithuanians to learn Lithuanian, to develop linguistic competence appropriate to their duties, and to adapt to a new environment. 8. RG is focused not only on setting out the basics of the language but also on reducing the distance between written and spoken language. 9. Since the description of the grammatical structure of the Lithuanian language in the mid-seventeenth century was suitable as a basis for a grammar of the mid-eighteenth century, it evidences the slight change in the language itself.

An Efficient FPGA-Based Welch Power Spectral Density for Real-Time Applications

Power spectral density is a crucial tool in the field of signal processing, mainly in biomedical signal processing. Power spectral density is also one of the most widely used tools in real-time applications. Therefore, further research should be prioritized in the hardware implementation of power spectral density. In this paper, two techniques are introduced in the implementation of power spectral density, mainly focusing on the Fourier transform block. The approaches introduced are the adoption of the Coordinate Rotation Digital Computer algorithm-based fast Fourier transform and the Coordinate Rotation Digital Computer algorithm-based sliding discrete Fourier transform. The other blocks in modified Welch’s architecture are also enhanced using pipelining and approximate distributed arithmetic methods. The introduction of all these techniques has led to an improvement in power and area. There is almost a 36% decrease in the number of lookup tables when compared to the existing methodology. With regards to power, there is almost 44% and 16% decrease in these two architectures, respectively. The overall architectures were synthesized using Xilinx Vivado 19.1, and the language used was Verilog Hardware Description Language.

The Classification and Language Description of Patients with Primary Progressive Aphasia Using the Mini Linguistic State Examination Test.

Introduction. Primary progressive aphasia (PPA) is a clinical syndrome characterized by a progressive deterioration in language and speech. It is classified into three variants based on symptom patterns: logopenic, semantic, and non-fluent. Due to the lack of fully reliable and valid screening tests for diagnosing PPA and its variants, a Spanish version of the Mini Linguistic State Examination (MLSE) has recently been introduced. Materials and methods. This study aimed to describe the language impairments in a small sample of six patients with PPA and classify them into the three variants using the decision tree and syndrome guide proposed by the MLSE authors. Results. The findings demonstrate the test's utility in classifying some PPA variants through a qualitative analysis of patient performance and error types. The study revealed a 50% accuracy rate for the decision tree and an 83.33% accuracy rate when using the syndrome guide. Discussion. This discrepancy arises because the decision tree often classified cases as logopenic variant PPA (lvPPA) when working memory was significantly impaired. Specifically, it tended to misclassify patients with semantic, motor, or speech impairments as having lvPPA due to its reliance on the sentence repetition task for assessing working memory.

Design and Analysis of Area and Time Efficient Signed Square Architecture Using Nonconventional Method

Computing the square of any large signed number using conventional methods is challenging. Therefore, parallel and partial product generation while squaring signed numbers is complex and time-consuming. This paper suggests a very large scale integration design for an efficient signed square architecture using a nonconventional method. The square of a large magnitude signed number is reduced to a lower magnitude signed multiplication and an addition operation by the Yavadunam algorithm. The suggested method finds the deficit of the signed number from the adjacent base to perform the signed square operation. The proposed design is coded using Verilog hardware description language. The synthesis and simulation of this architecture are performed using Xilinx integrated synthesis environment 14.7 software. Furthermore, this architecture is implemented on the Virtex-4, spartan-3 and spartan-6 field programmable gate array devices. The design is also coded in cadence virtuoso centos version 5. Then, it is synthesized using 180[Formula: see text]nm and 90[Formula: see text]nm technology. The validation of the proposed architecture is examined using the performance parameters, such as path delay and area. The synthesis results show improvements in delay-area complexity compared to state-of-the-art architectures. The superiority of the proposed architecture is claimed based on performance comparisons with the previous square, multipliers and signed multiplier architectures.

FPGA design and implementation for montgomery multiplication algorithm using MATLAB HDL coder

BackgroundModular multiplication for large numbers is especially important in cryptography algorithms such as RSA and elliptic curves. The Montgomery algorithm is the most famous and efficient one for calculating it. Hardware implementation for cryptography co-processors is better than software implementation in terms of speed and security. Many FPGA designs for the Montgomery multiplication algorithm was published based on hardware description languages like VERILOG and VHDL. This paper proposes the FPGA design and implementation using MATLAB HDL Coder.ResultsThe algorithm is modified such that it can fit any small/large FPGA by introducing scaling factor. The design is configurable in both modulus length and the scaling factor. This paper performs a comparison between the synthesizing results for different scales and for different modulus lengths. The synthesizing is performed up to 8K bit modulus length, and it can be increased easily. In this paper, implementation of different modulus lengths with different frequencies and with different area utilization can be easily achieved. The design utilizes different area resources for each configuration. The target is xc7vx330t-2ffg1157 Virtex-7 Xilinx FPGA. The maximum frequency is 80.81 MHz for 4096-bit modulus length with 8-bit data width and 2 for serialization factor. The minimum area utilization is achieved for minimum configurations, i.e., 1024-bit modulus length with 8-bit data width and for unity serialization factor.ConclusionsThis paper proposes a scalable and configurable FPGA design for Montgomery multiplication co-processor-based HDL coder design.

Efficient hardware implementation of ROM-free LFM waveform generation for real-time reconfigurable software-defined radar

Abstract This paper presents a novel method for generating Linear Frequency Modulation (LFM) waveforms. Conventional Look-Up Table (LUT)-based method require large storage resources, especially in high-resolution, multi-mode radar systems. To address this issue, we propose a ROM-free, real-time LFM waveform generator. It uses the fourth-order Runge-Kutta (RK-4) method to solve differential equations for the in-phase and quadrature components. The proposed approach dynamically generates LFM waveforms using only parameters such as pulse width, bandwidth, and sampling rate, eliminating the need for pre-calculated waveforms and LUTs. This method completely removes storage requirements, enabling faster synthesis times and achieving minimal latency. The generator’s performance is validated through simulation and FPGA implementation using a full hardware description language. Results confirm its effectiveness, making it a strong candidate for advanced radar systems that demand high performance with limited hardware resources.

Penrose Dodecahedron, Witting Configuration and Quantum Entanglement

A model with two entangled spin-3/2 particles based on geometry of dodecahedron was suggested by Roger Penrose for formulation of analogue of Bell theorem without probabilities. The model was later reformulated using so-called Witting configuration with 40 rays in 4D Hilbert space. However, such reformulation needs for some subtleties related with entanglement of two such configurations essential for consideration of non-locality and some other questions. Two entangled systems with quantum states described by Witting configurations are discussed in the present work. Duplication of points with respect to vertices of dodecahedron produces rather significant increase with number of symmetries in 25920/60=432 times. Quantum circuits model is a natural language for description of operations with different states and measurements of such systems.Quanta 2024; 13: 38–46.