Décio Luiz Gazzoni Filho, Gora Adj, Slim Bettaieb
et al.
Many post-quantum cryptosystems require generating an n-bit binary vector with a prescribed Hamming weight ω, a process known as fixed-weight sampling. When ω = O(n), we call this dense fixed-weight sampling, which commonly appears in lattice-based cryptosystems, like those in the NTRU family. In contrast, code-based cryptosystems typically use sparse fixed-weight sampling with ω = o(n) (e.g., O(√n). Sparse fixed-weight sampling generally involves three constant-time steps to keep the sampled vector secret: 1. sample ω nearly uniform random integers from a series of decreasing intervals; 2. map these integers into a set of ω distinct indices in [0, n), called the support; 3. generate a binary n-bit vector with bits set only at the support indices. Remarkably, some of the core algorithms employed in fixed-weight sampling date back to nearly a century, yet developing efficient and secure techniques remains essential for modern post-quantum cryptographic applications.
In this paper, we present novel algorithms for steps two and three of the fixedweight sampling process. We demonstrate their practical applicability by replacing the current fixed-weight sampling routine in the HQC post-quantum key exchange mechanism, recently selected for NIST standardization. We rigorously prove that our procedures are sound, secure, and introduce little to no bias. Our implementation of the proposed algorithms accelerates step 2 by up to 2.9x and step 3 by up to 5.8x compared to an optimized version of the fixed-weight sampler currently used in HQC. Since fixed-weight sampling constitutes a significant portion of HQC’s execution time, these speedups translate into protocol-level improvements of up to 1.37x, 1.28x and 1.21x for key generation, encapsulation and decapsulation, respectively.
Computer engineering. Computer hardware, Information technology
Paolo Dizon, Kyla Luisa P. Guanlao, Keona R. Marqueses
et al.
Concrete production uses large amounts of cement and natural aggregates. This makes it costly and harmful to the environment. This paper reviews accessible alternatives that offer lower costs while maintaining mechanical performance. These incorporate laterite aggregates, fly ash, sisal fibers, and superplasticizers. Laterite replaces a portion of the coarse aggregate, lowering the demand for natural stone. Fly ash is a partial cement replacement, which helps cut carbon emissions and diminish overall cost. Sisal fibers improve crack resistance and contribute to higher tensile and flexural strength. Superplasticizers improve workability without adding more water. These additives were combined and tested using different concrete blend designs. In the various gathered literature, it was found that one of the best-performing mixes had 10.52 % fly ash, 1 % sisal fiber, and 1.48 % superplasticizer. It reached 47.24 MPa compressive strength, 4.92 MPa tensile strength, and 5.21 MPa flexural strength with a lower production cost. The paper also considered other additives such as ground calcium carbonate, silica fume, and styrene-butadiene rubber, which have been linked to enhancing the mechanical properties.
Chemical engineering, Computer engineering. Computer hardware
Manuel Benavent-Lledo, David Mulero-Pérez, David Ortiz-Perez
et al.
Detecting actions as they occur is essential for applications like video surveillance, autonomous driving, and human-robot interaction. Known as online action detection, this task requires classifying actions in streaming videos, handling background noise, and coping with incomplete actions. Transformer architectures are the current state-of-the-art, yet the potential of recent advancements in computer vision, particularly vision-language models (VLMs), remains largely untapped for this problem, partly due to high computational costs. In this paper, we introduce TOAD: a Text-driven Online Action Detection architecture that supports zero-shot and few-shot learning. TOAD leverages CLIP (Contrastive Language-Image Pretraining) textual embeddings, enabling efficient use of VLMs without significant computational overhead. Our model achieves 82.46% mAP on the THUMOS14 dataset, outperforming existing methods, and sets new baselines for zero-shot and few-shot performance on the THUMOS14 and TVSeries datasets.
Transformers have proven effective in language modeling but are limited by high computational and memory demands that grow quadratically with input sequence length. State space models (SSMs) offer a promising alternative by reducing attention complexity from $O(L^2)$ to $O(L)$ while also lowering overall memory consumption. Vision Mamba adapts the SSM approach for computer vision tasks, achieving lower latency and memory consumption than traditional transformer models. However, deploying Vision Mamba on edge devices is challenging due to its sequential scan operations, which hinder GPU efficiency. We propose Mamba-X, an end-to-end Vision Mamba accelerator that includes a systolic scan array to maximize parallelism and minimize memory traffic, along with a hybrid, hardware-friendly quantization technique to reduce memory usage and improve hardware efficiency without sacrificing accuracy.
Abdelmalek Bouguettaya, Hafed Zarzour, A. Kechida
et al.
Vehicle detection from unmanned aerial vehicle (UAV) imagery is one of the most important tasks in a large number of computer vision-based applications. This crucial task needed to be done with high accuracy and speed. However, it is a very challenging task due to many characteristics related to the aerial images and the used hardware, such as different vehicle sizes, orientations, types, density, limited datasets, and inference speed. In recent years, many classical and deep-learning-based methods have been proposed in the literature to address these problems. Handed engineering- and shallow learning-based techniques suffer from poor accuracy and generalization to other complex cases. Deep-learning-based vehicle detection algorithms achieved better results due to their powerful learning ability. In this article, we provide a review on vehicle detection from UAV imagery using deep learning techniques. We start by presenting the different types of deep learning architectures, such as convolutional neural networks, recurrent neural networks, autoencoders, generative adversarial networks, and their contribution to improve the vehicle detection task. Then, we focus on investigating the different vehicle detection methods, datasets, and the encountered challenges all along with the suggested solutions. Finally, we summarize and compare the techniques used to improve vehicle detection from UAV-based images, which could be a useful aid to researchers and developers to select the most adequate method for their needs.
In this research paper, main concentrate of the authors on the profit comparison of computer system with hardware redundancy by introducing the concept of priority to software up-gradation, hardware preventive maintenance (PM) and hardware maximum repair time (MRT). The system fails independently from normal mode. All the repair activities such as hardware repair, software up-gradation, hardware preventive maintenance before failure and hardware replacement after maximum repair time are carried out by a single server immediately on need basis. All random variables are statistically independent. The negative exponential distribution is taken for the failure time of the component while the distributions of repair time, up-gradation time, preventive maintenance and replacement time are assumed arbitrary with different probability density functions. Semi-Markov process and regenerative point technique are used. The behaviour of profits of the system models have been examined for different parameters and costs.
Fachri Amsury, Nanang Ruhyana, Andri Agung Riyadi
et al.
Manufacturing companies rely on machines for operational activities to produce finished goods. Common factors constraining the demand and supply of spare parts are the high number of spare parts managed and irregular patterns of demand for spare parts. These varying quantities also require investment in spare parts inventory and longer response times than predicted. The research aims to apply the FP-Growth algorithm approach to find association rules and produce patterns of demand and supply of spare parts in lightweight brick manufacturing companies based on transaction data on demand and supply of spare parts from January – March 2023. The approach used is associated with the applied algorithm. In this research, the primary process of the FP-Growth algorithm is to create a combination of each item until no more combinations are formed using minimum support and minimum confidence parameters. Based on the results of making association rules using spare parts demand data from the machine maintenance department, it is stated that the regulations formed from processing the RapidMiner application with a confidence value of 100% recommend FD Regular Bolt spare parts, then the next rating with a confidence value of 94% is Steel Nuts, seven rules recommend Nuts. Steel. Therefore, it is recommended that FD Regular Bolts and Steel Nuts carry out safety stock to maintain stock availability and place them on shelves included in the fast-moving inventory category.
Brandon Warner, Edward Ratner, Kallin Carlous-Khan
et al.
This paper proposes a novel model-agnostic method for weighting the outputs of base classifiers in machine learning (ML) ensembles. Our approach uses class-based weight coefficients assigned to every output class in each learner in the ensemble. This is particularly useful when the base classifiers have highly variable performance across classes. Our method generates a dense set of coefficients for the models in our ensemble by considering the model performance on each class. We compare our novel method to the commonly used ensemble approaches like voting and weighted averages. In addition, we compare our approach to class-specific soft voting (CSSV), which was also designed to address variable performance but generates a sparse set of weights by solving a linear system. We choose to illustrate the power of this approach by applying it to an ensemble of extreme learning machines (ELMs), which are well suited for this approach due to their stochastic, highly varying performance across classes. We illustrate the superiority of our approach by comparing its performance to that of simple majority voting, weighted majority voting, and class-specific soft voting using ten popular open-source multiclass classification datasets.
Commercialised biodiesel, comprising of methyl esters, have large amounts of oxygenated compounds which cause low calorific value, high viscosity, poor low temperature performance and can only used in blends with petroleum-based diesel. Deoxygenation of these compounds can occur through three reaction pathways which are decarboxylation, decarbonylation, hydrodeoxygenation. The removal of oxygen can improve their fuel properties, and is heavily influenced by the presence of hydrogen (H2). However, given safety issues during transportation and storage of H2, the use of solvents to produce in-situ H2 for the deoxygenation process has been suggested as an alternative. The present work attempts to investigate the addition of solvents (deionised water, methanol, ethanol, 1-propanol, 2-propanol, n-hexane and cyclohexane) on the deoxygenation of methyl oleate to enhance its fuel properties. The experiments were carried out with unreduced bimetallic NiCo impregnated onto TiO2. Incorporation of NiCo onto TiO2 maintained the mesoporous nature of the support while increasing the number of strong acid sites of the catalyst, which can promote C-O bond cleavage during deoxygenation. Under hydrogen-restricted conditions, the deoxygenation is expected to occur mainly through decarboxylation and decarbonylation to produce alkanes and alkenes, with cracking to produce shorter chain methyl esters. The deoxygenation was conducted with 50 g methyl oleate, 40 g solvent, 5 wt% catalyst at 300 °C for 2 h in a pressurised nitrogen atmosphere. GCMS analysis show that the addition of 2-propanol showed the highest methyl oleate conversion (69.08 %) over other solvents, with 12.74 % selectivity for alkane formation. These results indicate the potential of solvent-aided deoxygenation of methyl esters for biofuel use.
Chemical engineering, Computer engineering. Computer hardware
This paper aims to verify the effectiveness of the software implementation of the proposed algorithm in creating computer-generated images of sand mixtures using a photograph of sand as an input and its effectiveness in converting digital pictures into sand-based images out of the mixtures it generated. The method of this paper is to visually compare the photographed image of the actual mixtures to its computer-generated counterpart to verify if the mixture generation produces results as expected and compare the computer-generated sand-based images with its source to verify image reproduction maintains same image content. The results of the mixture comparison shows that the actual and the computer-generated ones have similar overall shade and color. Still, the generated one has a rougher texture and higher contrast due to the method of inheriting visual features by pixel, not by individual sand particles. The comparison of the sand-based image and its source has demonstrated the software's ability to maintain the essence of its contents during conversion while replacing its texture with the visual properties of the generated sand mixture. The result have shown that the software implementation of the proposed algorithm can effectively use the images of sand to generate images of its mixtures and use those mixture images to convert a digital picture into a computer-generated sand-based image.
A paradigm shift is underway in Software Engineering, with AI systems such as LLMs playing an increasingly important role in boosting software development productivity. This trend is anticipated to persist. In the next years, we expect a growing symbiotic partnership between human software developers and AI. The Software Engineering research community cannot afford to overlook this trend; we must address the key research challenges posed by the integration of AI into the software development process. In this paper, we present our vision of the future of software development in an AI-driven world and explore the key challenges that our research community should address to realize this vision.
Recent breakthroughs in quantum hardware are creating opportunities for its use in many applications. However, quantum software engineering is still in its infancy with many challenges, especially dealing with the diversity of quantum programming languages and hardware platforms. To alleviate these challenges, we propose QFaaS, a novel Quantum Function-as-a-Service framework, which leverages the advantages of the serverless model and the state-of-the-art software engineering approaches to advance practical quantum computing. Our framework provides essential components of a quantum serverless platform to simplify the software development and adapt to the quantum cloud computing paradigm, such as combining hybrid quantum-classical computation, containerizing functions, and integrating DevOps features. We design QFaaS as a unified quantum computing framework by supporting well-known quantum languages and software development kits (Qiskit, Q#, Cirq, and Braket), executing the quantum tasks on multiple simulators and quantum cloud providers (IBM Quantum and Amazon Braket). This paper proposes architectural design, principal components, the life cycle of hybrid quantum-classical function, operation workflow, and implementation of QFaaS. We present two practical use cases and perform the evaluations on quantum computers and simulators to demonstrate our framework's ability to ease the burden on traditional engineers to expedite the ongoing quantum software transition.
Quantum computing has potential to provide exponential speedups over classical computing for many important applications. However, today's quantum computers are in their early stages, and hardware quality issues hinder the scale of program execution. Benchmarking and simulation of quantum circuits on classical computers is therefore essential to advance the understanding of how quantum computers and programs operate, enabling both algorithm discovery that leads to high-impact quantum computation and engineering improvements that deliver to more powerful quantum systems. Unfortunately, the nature of quantum information causes simulation complexity to scale exponentially with problem size. In this paper, we debut Super.tech's SuperSim framework, a new approach for high fidelity and scalable quantum circuit simulation. SuperSim employs two key techniques for accelerated quantum circuit simulation: Clifford-based simulation and circuit cutting. Through the isolation of Clifford subcircuit fragments within a larger non-Clifford circuit, resource-efficient Clifford simulation can be invoked, leading to significant reductions in runtime. After fragments are independently executed, circuit cutting and recombination procedures allow the final output of the original circuit to be reconstructed from fragment execution results. Through the combination of these two state-of-art techniques, SuperSim is a product for quantum practitioners that allows quantum circuit evaluation to scale beyond the frontiers of current simulators. Our results show that Clifford-based circuit cutting accelerates the simulation of near-Clifford circuits, allowing 100s of qubits to be evaluated with modest runtimes.
Abstract Tons of products are facilitated through powder mixing processes. The quality of the final products is highly influenced by the mixing state. Hence, detailed information about the mixing state is necessary to improve the quality. Recent remarkable advancements in computer hardware enable the evaluation of the mixing state via numerical simulation. The Discrete Element Method (DEM) is frequently used in numerical simulations, where computational particle shapes are typically modeled by a spherical body. On the other hand, not all particle shapes in industries are spherical, and the structure of industrial mixers is generally complex. Thus, when the DEM is applied to an industrial mixer, modeling of particle shape and the complex structure of the mixers may be taken into consideration. In the present study, an innovative model is newly developed to simulate the powder mixing of non-spherical particles in an industrial mixer. Here, a non-spherical particle is modeled by the ellipsoidal equation, and an industrial mixer is modeled using the Signed Distance Function (SDF), where the wall boundary shape is flexibly modeled by a scalar field. This approach is referred to as the ellipsoidal DEM/SDF. In the current study, the adequacy and applicability of the ellipsoidal DEM/SDF are demonstrated by the following steps. First, the compatibility of the ellipsoidal DEM with the SDF wall boundary model and the applicability of a soft linear spring in the ellipsoidal DEM/SDF are demonstrated through validation tests. Subsequently, the effect of particle shape on mixing progress in a ribbon mixer is examined. Through computations, the effect of the non-spherical shape is shown to be insignificant on powder mixing progress in the ribbon mixer. Consequently, the ellipsoidal DEM/SDF is shown to be applicable to the industrial powder mixing, and therefore, will contribute to optimization of the mixer design and the operational conditions in chemical, food, and pharmaceutical engineering fields.
Simulation-based education (SE) refers to the use of simulation software, tools, and serious games to enrich the teaching and learning processes. Advances in both computer hardware and software allow for employing innovative methodologies that make use of SE tools to enhance the learning experience. Moreover, thanks to the globalisation of e-learning practices, these educational experiences can be made available to students from different geographical regions and universities, which promotes the development of international and inter-university cooperation in education. This paper provides a review of recent works in the SE subject, with a focus on the areas of engineering, science, and management. It also discusses some experiences in SE involving different European universities and learning models. Finally, it also points out open challenges as well as noticeable trends.
Muhammad Daffa Arviano Putra, Muhammad Darwis, Retno Hendrowati
Forum adalah suatu wadah untuk bertukar informasi dan pemikiran. Dalam era digital, informasi merupakan hal yang paling berharga dan forum menjadi sarana tepat untuk mendapatkan informasi. Akibatnya, forum saat ini juga hadir dalam bentuk aplikasi. Suatu aplikasi forum dapat memberikan banyak manfaat bagi penggunanya. Pada penelitian ini, dirancang aplikasi forum khusus mahasiswa Universitas Paramadina berbentuk mobile. Kehadiran aplikasi forum mahasiswa ini diharapkan meningkatkan kolaborasi dan semangat diskusi di antara mahasiswa, sehingga memperkaya pengalaman belajar mahasiswa di Universitas Paramadina. Proses pengembangan sistem aplikasi forum dilakukan dengan menerapkan konsep perancangan berbasis objek. Konsep ini diterapkan dengan pembuatan model-model diagram Unified Modeling Language (UML). Dengan demikian, rancangan sistem yang dibuat mengikuti struktur yang efisien dan mengikuti standard. Hasil rancangan kemudian diimplementasikan dalam bentuk prototype dengan memanfaatkan framework Flutter dan dilakukan pengujian secara menyeluruh untuk memastikan ketepatan dan kegunaan aplikasi dengan menggunakan metode blackbox dan whitebox. Hasil pengujian tersebut menunjukkan bahwa aplikasi forum tersebut telah berfungsi dengan baik.
Zheyuan Hu, Zekun Shi, George Em Karniadakis
et al.
Physics-Informed Neural Networks (PINNs) have proven effective in solving partial differential equations (PDEs), especially when some data are available by seamlessly blending data and physics. However, extending PINNs to high-dimensional and even high-order PDEs encounters significant challenges due to the computational cost associated with automatic differentiation in the residual loss. Herein, we address the limitations of PINNs in handling high-dimensional and high-order PDEs by introducing Hutchinson Trace Estimation (HTE). Starting with the second-order high-dimensional PDEs ubiquitous in scientific computing, HTE transforms the calculation of the entire Hessian matrix into a Hessian vector product (HVP). This approach alleviates the computational bottleneck via Taylor-mode automatic differentiation and significantly reduces memory consumption from the Hessian matrix to HVP. We further showcase HTE's convergence to the original PINN loss and its unbiased behavior under specific conditions. Comparisons with Stochastic Dimension Gradient Descent (SDGD) highlight the distinct advantages of HTE, particularly in scenarios with significant variance among dimensions. We further extend HTE to higher-order and higher-dimensional PDEs, specifically addressing the biharmonic equation. By employing tensor-vector products (TVP), HTE efficiently computes the colossal tensor associated with the fourth-order high-dimensional biharmonic equation, saving memory and enabling rapid computation. The effectiveness of HTE is illustrated through experimental setups, demonstrating comparable convergence rates with SDGD under memory and speed constraints. Additionally, HTE proves valuable in accelerating the Gradient-Enhanced PINN (gPINN) version as well as the Biharmonic equation. Overall, HTE opens up a new capability in scientific machine learning for tackling high-order and high-dimensional PDEs.