Performance Gain Research Articles

CuGBasis: High-performance CUDA/Python library for efficient computation of quantum chemistry density-based descriptors for larger systems.

CuGBasis is a free and open-source CUDA®/Python library for efficient computation of scalar, vector, and matrix quantities crucial for the post-processing of electronic structure calculations. CuGBasis integrates high-performance Graphical Processing Unit (GPU) computing with the ease and flexibility of Python programming, making it compatible with a vast ecosystem of libraries. We showcase its utility as a Python library and demonstrate its seamless interoperability with existing Python software to gain chemical insight from quantum chemistry calculations. Leveraging GPU-accelerated code, cuGBasis exhibits remarkable performance, making it highly applicable to larger systems or large databases. Our benchmarks reveal a 100-fold performance gain compared to alternative software packages, including serial/multi-threaded Central Processing Unit and GPU implementations. This paper outlines various features and computational strategies that lead to cuGBasis's enhanced performance, guiding developers of GPU-accelerated code.

FedCCL: Federated dual-clustered feature contrast under domain heterogeneity

Federated learning (FL) facilitates a privacy-preserving neural network training paradigm through collaboration between edge clients and a central server. One significant challenge is that the distributed data is not independently and identically distributed (non-IID), typically including both intra-domain and inter-domain heterogeneity. However, recent research is limited to simply using averaged signals as a form of regularization and only focusing on one aspect of these non-IID challenges. Given these limitations, this paper clarifies these two non-IID challenges and attempts to introduce cluster representation to address them from both local and global perspectives. Specifically, we propose a dual-clustered feature contrast-based FL framework with dual focuses. First, we employ clustering on the local representations of each client, aiming to capture intra-class information based on these local clusters at a high level of granularity. Then, we facilitate cross-client knowledge sharing by pulling the local representation closer to clusters shared by clients with similar semantics while pushing them away from clusters with dissimilar semantics. Second, since the sizes of local clusters belonging to the same class may differ for each client, we further utilize clustering on the global side and conduct averaging to create a consistent global signal for guiding each local training in a contrastive manner. Experimental results on multiple datasets demonstrate that our proposal achieves comparable or superior performance gain under intra-domain and inter-domain heterogeneity.

Regularization-based methods for ordinal quantification

Quantification, i.e., the task of predicting the class prevalence values in bags of unlabeled data items, has received increased attention in recent years. However, most quantification research has concentrated on developing algorithms for binary and multi-class problems in which the classes are not ordered. Here, we study the ordinal case, i.e., the case in which a total order is defined on the set of n>2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n>2$$\\end{document} classes. We give three main contributions to this field. First, we create and make available two datasets for ordinal quantification (OQ) research that overcome the inadequacies of the previously available ones. Second, we experimentally compare the most important OQ algorithms proposed in the literature so far. To this end, we bring together algorithms proposed by authors from very different research fields, such as data mining and astrophysics, who were unaware of each others’ developments. Third, we propose a novel class of regularized OQ algorithms, which outperforms existing algorithms in our experiments. The key to this gain in performance is that our regularization prevents ordinally implausible estimates, assuming that ordinal distributions tend to be smooth in practice. We informally verify this assumption for several real-world applications.

Deep Learning for Point-of-Care Ultrasound Image Quality Enhancement: A Review

The popularity of handheld devices for point-of-care ultrasound (POCUS) has increased in recent years due to their portability and cost-effectiveness. However, POCUS has the drawback of lower imaging quality compared to conventional ultrasound because of hardware limitations. Improving the quality of POCUS through post-image processing would therefore be beneficial, with deep learning approaches showing promise in this regard. This review investigates the state-of-the-art progress of image enhancement using deep learning suitable for POCUS applications. A systematic search was conducted from January 2024 to February 2024 on PubMed and Scopus. From the 457 articles that were found, the full text was retrieved for 69 articles. From this selection, 15 articles were identified addressing multiple quality enhancement aspects. A disparity in the baseline performance of the low-quality input images was seen across these studies, ranging between 8.65 and 29.24 dB for the Peak Signal-to-Noise Ratio (PSNR) and between 0.03 an 0.71 for the Structural Similarity Index Measure (SSIM). In six studies, where both the PSNR and the SSIM metrics were reported for the baseline and the generated images, mean differences of 6.60 (SD ± 2.99) and 0.28 (SD ± 0.15) were observed for the PSNR and SSIM, respectively. The reported performance outcomes demonstrate the potential of deep learning-based image enhancement for POCUS. However, variability in the extent of the performance gain across datasets and articles was notable, and the heterogeneity across articles makes quantifying the exact improvements challenging.

Extended version—to be, or not to be stateful: post-quantum secure boot using hash-based signatures

AbstractWhile research in PQC has gained significant momentum, its adoption in real-world products is slow. This is largely due to concerns about practicability and maturity. The secure boot process of embedded devices is one scenario where such restraints can result in fundamental security problems. In this work, we present a flexible hardware/software co-design for HBS schemes which enables the transition to a post-quantum secure boot today. These signature schemes stand out due to their straightforward security proofs and are on the fast track to standardisation. Unlike previous work, we exploit the performance intensive similarities of the stateful LMS and XMSS schemes as well as the stateless $$\text {SPHINCS}^{+}$$ SPHINCS + scheme. Thus, we enable designers to use a stateful or stateless scheme depending on the constraints of each individual application. To demonstrate the feasibility of our approach, we compare our results with hardware accelerated implementations of classical asymmetric algorithms. Further, we outline the use of different HBS schemes during the boot process. We compare different schemes, show the importance of parameter choices, and demonstrate the performance gain with different levels of hardware acceleration.

Efficient chemical equilibria calculation by artificial neural networks for ammonia cracking and synthesis

The calculation of chemical equilibria in detailed reactor simulations frequently requires elaborate numerical solution of the governing equations in an iterative way, which is often computationally expensive and can significantly increase the overall computation time. In order to reduce these computational costs, we introduce a ready-to-use tool, ANNH3, for calculation of equilibrium composition for synthesis and cracking of ammonia based on a neural network. This tool provides excellent agreement with the conventional approach in the range of 135–1000 °C and 1–100 bar and is ca. 100 times faster than conventional stoichiometry-based concepts by replacing the iterative solution process with neural network inference. While speed-up is significant even for the relatively simple case of ammonia synthesis and decomposition, we expect an even higher performance gain for the equilibrium calculation in reaction systems where more components and multiple reactions are involved.

Interpolation-split: a data-centric deep learning approach with big interpolated data to boost airway segmentation performance

The morphology and distribution of airway tree abnormalities enable diagnosis and disease characterisation across a variety of chronic respiratory conditions. In this regard, airway segmentation plays a critical role in the production of the outline of the entire airway tree to enable estimation of disease extent and severity. Furthermore, the segmentation of a complete airway tree is challenging as the intensity, scale/size and shape of airway segments and their walls change across generations. The existing classical techniques either provide an undersegmented or oversegmented airway tree, and manual intervention is required for optimal airway tree segmentation. The recent development of deep learning methods provides a fully automatic way of segmenting airway trees; however, these methods usually require high GPU memory usage and are difficult to implement in low computational resource environments. Therefore, in this study, we propose a data-centric deep learning technique with big interpolated data, Interpolation-Split, to boost the segmentation performance of the airway tree. The proposed technique utilises interpolation and image split to improve data usefulness and quality. Then, an ensemble learning strategy is implemented to aggregate the segmented airway segments at different scales. In terms of average segmentation performance (dice similarity coefficient, DSC), our method (A) achieves 90.55%, 89.52%, and 85.80%; (B) outperforms the baseline models by 2.89%, 3.86%, and 3.87% on average; and (C) produces maximum segmentation performance gain by 14.11%, 9.28%, and 12.70% for individual cases when (1) nnU-Net with instant normalisation and leaky ReLU; (2) nnU-Net with batch normalisation and ReLU; and (3) modified dilated U-Net are used respectively. Our proposed method outperformed the state-of-the-art airway segmentation approaches. Furthermore, our proposed technique has low RAM and GPU memory usage, and it is GPU memory-efficient and highly flexible, enabling it to be deployed on any 2D deep learning model.

An Ensemble Spectral Prediction (ESP) model for metabolite annotation.

A key challenge in metabolomics is annotating measured spectra from a biological sample with chemical identities. Currently, only a small fraction of measurements can be assigned identities. Two complementary computational approaches have emerged to address the annotation problem: mapping candidate molecules to spectra, and mapping query spectra to molecular candidates. In essence, the candidate molecule with the spectrum that best explains the query spectrum is recommended as the target molecule. Despite candidate ranking being fundamental in both approaches, limited prior works incorporated rank learning tasks in determining the target molecule. We propose a novel machine learning model, Ensemble Spectral Prediction (ESP), for metabolite annotation. ESP takes advantage of prior neural network-based annotation models that utilize multilayer perceptron (MLP) networks and Graph Neural Networks (GNNs). Based on the ranking results of the MLP- and GNN-based models, ESP learns a weighting for the outputs of MLP and GNN spectral predictors to generate a spectral prediction for a query molecule. Importantly, training data is stratified by molecular formula to provide candidate sets during model training. Further, baseline MLP and GNN models are enhanced by considering peak dependencies through label mixing and multi-tasking on spectral topic distributions. When trained on the NIST 2020 dataset and evaluated on the relevant candidate sets from PubChem, ESP improves average rank by 23.7% and 37.2% over the MLP and GNN baselines, respectively, demonstrating performance gain over state-of-the-art neural network approaches. However, MLP approaches remain strong contenders when considering top five ranks. Importantly, we show that annotation performance is dependent on the training dataset, the number of molecules in the candidate set and candidate similarity to the target molecule. The ESP code, a trained model, and a Jupyter notebook that guide users on using the ESP tool is available at https://github.com/HassounLab/ESP.

Coupled Conditional Neural Movement Primitives

Learning sensorimotor trajectories through flexible neural representations is fundamental for robots as it facilitates the building of motor skills as well as equipping them with the ability to represent the world as predictable temporal events. Recent advances in deep learning led to the development of powerful learning from demonstration (LfD) systems such as Conditional Neural Movement Primitives (CNMPs). CNMPs can robustly represent skills as movement distributions and allow them to be ‘recalled’ by conditioning the movement on a few observation points. In this study, we focus on improving CNMPs to achieve a higher resource economy by adopting a divide-and-conquer approach. We propose a novel neural architecture called Coupled CNMP (C-CNMP), that couples the latent spaces of a pair of CNMPs that splits a given sensorimotor trajectory into segments whose learning is undertaken by smaller sub-networks. Therefore, each sub-network needs to deal with a less complex trajectory making the learning less resource-hungry. With systematic simulations on a controlled trajectory data set, we show that the overhead brought by the coupling introduced in our model is well offset by the resource and performance gain obtained. To be specific, with CNMP model as the baseline, it is shown that the proposed model is able to learn to generate trajectories in the data set with a lower trajectory error measured as the mean absolute difference between the generated trajectory and the ground truth. Importantly, our model can perform well with relatively limited resources, i.e., with less number of neural network parameters compared to the baseline. To show that the findings from the controlled data set well-transfer to robot data, we use robot joint data in an LfD setting and compare the learning performance of the proposed model with the baseline model at equal complexity levels. The simulation experiments show that with also the robot joint data, the proposed model, C-CNMP, learns to generate the joint trajectories with significantly less error than the baseline model. Overall, our study improves the state of the art in sensorimotor trajectory learning and exemplifies how divide-and-conquer approaches can benefit deep learning architectures for resource economy.

On the emerging potential of quantum annealing hardware for combinatorial optimization

Over the past decade, the usefulness of quantum annealing hardware for combinatorial optimization has been the subject of much debate. Thus far, experimental benchmarking studies have indicated that quantum annealing hardware does not provide an irrefutable performance gain over state-of-the-art optimization methods. However, as this hardware continues to evolve, each new iteration brings improved performance and warrants further benchmarking. To that end, this work conducts an optimization performance assessment of D-Wave Systems’ Advantage Performance Update computer, which can natively solve sparse unconstrained quadratic optimization problems with over 5,000 binary decision variables and 40,000 quadratic terms. We demonstrate that classes of contrived problems exist where this quantum annealer can provide run time benefits over a collection of established classical solution methods that represent the current state-of-the-art for benchmarking quantum annealing hardware. Although this work does not present strong evidence of an irrefutable performance benefit for this emerging optimization technology, it does exhibit encouraging progress, signaling the potential impacts on practical optimization tasks in the future.

Investigation of double-PCM based PV composite wall for power-generation and building insulation: Thermal characteristics and energy consumption prediction

The integration of phase change material (PCM) with building-integrated photovoltaic (BIPV) presents a compelling approach to enhance solar energy utilization and mitigate indoor thermal loads, contributing to energy-efficient and low-carbon building development. Traditional BIPV-PCM structures, however, struggle to balance PV efficiency and thermal insulation, particularly with varying PCM wall positions. To address this situation, this study introduces a novel double-PCM BIPV composite envelope (BIPV-dPCM). An experimentally validated dynamic heat transfer model was developed and used to perform a comparative simulation analysis with three reference systems to quantify the energy-saving potential of the BIPV-dPCM, focusing on PV output and wall insulation effectiveness metrics. Further dimensionless parametric analysis were carried out to investigate the systematic performance of the two PCMs at different relativities. In addition, the coupled working mechanism of the BIPV-dPCM system concerning the power generation performance and thermal insulation performance under transient variations is explored. It was found that the BIPV-dPCM showcases superior thermoelectric coupling performance compared to three alternative enclosures. Incorporating two PCMs significantly enhances electrical exergy efficiency by 11.66 % and thermal exergy efficiency by 1.54 %, surpassing other reference systems. The increase in PCM latent heat ratio has a limited effect on performance gain. Notably, as the PCM thickness ratio exceeds 1, the decline in P value decelerates, for every 0.5 increment in the g, the P value diminishes by merely 0.2 %. The ideal h is identified between 1 and 1.5, with 1.5 being optimal for energy conservation objectives. Additionally, the self-sufficiency coefficient (SSC) of the BIPV-dPCM remains robust, sustaining a range of 55 % to 65 % over prolonged periods. This study offers novel perspectives and serves as a design reference for optimizing building energy systems and enhancing cooling efficiencies in subtropical climates.

Redefining Neural Architecture Search of Heterogeneous Multinetwork Models by Characterizing Variation Operators and Model Components.

With neural architecture search (NAS) methods gaining ground on manually designed deep neural networks-even more rapidly as model sophistication escalates-the research trend is shifting toward arranging different and often increasingly complex NAS spaces. In this conjuncture, delineating algorithms which can efficiently explore these search spaces can result in a significant improvement over currently used methods, which, in general, randomly select the structural variation operator, hoping for a performance gain. In this article, we investigate the effect of different variation operators in a complex domain, that of multinetwork heterogeneous neural models. These models have an extensive and complex search space of structures as they require multiple subnetworks within the general model in order to answer different output types. From that investigation, we extract a set of general guidelines whose application is not limited to that particular type of model and are useful to determine the direction in which an architecture optimization method could find the largest improvement. To deduce the set of guidelines, we characterize both the variation operators, according to their effect on the complexity and performance of the model; and the models, relying on diverse metrics which estimate the quality of the different parts composing it.

Task-Related Saliency for Few-Shot Image Classification.

A weakness of the existing metric-based few-shot classification method is that task-unrelated objects or backgrounds may mislead the model since the small number of samples in the support set is insufficient to reveal the task-related targets. An essential cue of human wisdom in the few-shot classification task is that they can recognize the task-related targets by a glimpse of support images without being distracted by task-unrelated things. Thus, we propose to explicitly learn task-related saliency features and make use of them in the metric-based few-shot learning schema. We divide the tackling of the task into three phases, namely, the modeling, the analyzing, and the matching. In the modeling phase, we introduce a saliency sensitive module (SSM), which is an inexact supervision task jointly trained with a standard multiclass classification task. SSM not only enhances the fine-grained representation of feature embedding but also can locate the task-related saliency features. Meanwhile, we propose a self-training-based task-related saliency network (TRSN) which is a lightweight network to distill task-related salience produced by SSM. In the analyzing phase, we freeze TRSN and use it to handle novel tasks. TRSN extracts task-relevant features while suppressing the disturbing task-unrelated features. We, therefore, can discriminate samples accurately in the matching phase by strengthening the task-related features. We conduct extensive experiments on five-way 1-shot and 5-shot settings to evaluate the proposed method. Results show that our method achieves a consistent performance gain on benchmarks and achieves the state-of-the-art.

Mesoscopic V2X simulation framework to enhance simulation performance

The rapid evolution of vehicular communication has led to numerous new algorithms and applications based on this technology. Neglecting issues arising from wireless communication, such as the loss of information and delays, can result in problems such as reduced performance or compromised safety. However, while simulating V2X demands significant computational resources, it proves unsuitable for complex testing setups, including mixed-reality testing. This paper enhances V2X simulation by relying on an ecosystem based on SUMO, OMNeT++, Veins, and INET simulation tools. The proposed novel method introduces mesoscopic simulation in Vehicular Ad-hoc Networks to increase simulation performance to a level where real-time behavior is achievable. Meanwhile, it can also be beneficial in the acceleration of regular simulations. The presented solution introduces Meso nodes that are capable of aggregating communication across an entire traffic area, facilitated by a neural network function approximator. Results showed substantial performance gain while simulation accuracy was preserved.

Discovering an inference recipe for weakly-supervised object localization

Most existing weakly-supervised object localization (WSOL) methods have improved training procedures for better localization performance. However, the inference procedure has been overlooked. We observe that the useful information for localization is missed by the current inference practice of WSOL. To address this limitation, we propose a new test-time ingredient for WSOL: binarizing the penultimate feature map and their corresponding weights of the last linear layer. With this simple remedy, the proposed method consistently improves the localization performance of the existing training methods for WSOL. Extensive evaluation including with three different backbone networks on three different WSOL benchmarks validates its effectiveness. In addition, we demonstrate our method is also able to improve weakly-supervised semantic segmentation performances on PASCAL VOC dataset. Lastly, since our method is only applied during the testing phase, our performance gain comes with negligible computational overheads.

The effect of task processing management on energy consumption at the edge of Internet of things network with using reinforcement learning method

In this paper, the task processing approach at the edge of the Internet of Things (IoT) network in improving energy consumption was investigated. First, the subject of maximizing the use of energy is defined as the main problem and then based on the state of assignment of task processing from end devices at the edge of the network to smart gateways and local servers at the edge using possible low bandwidth and low battery consumption, the problem is divided into two subject which are obtained independently. After modeling by recombining these two sub-problems, the subject of maximizing the utility gain and battery life of the end devices is solved despite the limitations of processing time and energy resources.Considering that the environment of the problem and how the terminal devices are connected with the edge devices are unknown, a repetitive reinforcement learning algorithm has been used to create an optimal solution to maximize the operational benefit of the energy consumed in edge processing. The results achieved from the simulation show the increase in network load and processing overhead with the increase in the number of active devices. The suggested method, while increasing the edge processing speed, decreases the delay and maximizes the performance gain of the IOT edge system compared to the central cloud system. It should be noted that when the number of active end devices dramatically increases, the energy consumption and bandwidth at the network edge improves, and energy consumption at the network edge decreases lonely by 12.5%.

Transformative Noise Reduction: Leveraging a Transformer-Based Deep Network for Medical Image Denoising

Medical image denoising has numerous real-world applications. Despite their widespread use, existing medical image denoising methods fail to address complex noise patterns and typically generate artifacts in numerous cases. This paper proposes a novel medical image denoising method that learns denoising using an end-to-end learning strategy. Furthermore, the proposed model introduces a novel deep–wider residual block to capture long-distance pixel dependencies for medical image denoising. Additionally, this study proposes leveraging multi-head attention-guided image reconstruction to effectively denoise medical images. Experimental results illustrate that the proposed method outperforms existing qualitative and quantitative evaluation methods for numerous medical image modalities. The proposed method can outperform state-of-the-art models for various medical image modalities. It illustrates a significant performance gain over its counterparts, with a cumulative PSNR score of 8.79 dB. The proposed method can also denoise noisy real-world medical images and improve clinical application performance such as abnormality detection.

Transferrable Subject-Independent Feature Representation for Discriminating EEG-Based Brain Signals

Subject-independent Electroencephalography (EEG) recognition remains challenging due to inherent variability of brain anatomy across different subjects. Such variability is further complicated by the Volume Conduction Effect (VCE) that introduces channel-interference noise, exacerbating subject-specific biases in the recorded EEG signals. Existing studies, often relying large datasets and entangled spatial-temporal features, struggle to overcome this bias, particularly in scenarios with limited EEG data. To this end, we propose a Temporal-Connective EEG Representation Learning (TCERL) framework that disentangles temporal and spatial feature learning. TCERL first employs an one-dimensional convolutional network to extract channel-specific temporal features, mitigating channel-interference noise caused by VCE. Building upon these temporal features, TCERL then leverages Graph Neural Networks to extract subject-invariant topological features from a functional brain network, constructed using the channel-specific features as nodes and functional connectivity as the adjacency matrix. This approach allows TCERL to capture robust representations of brain activity patterns that generalize well across different subjects. Our empirical experiment demonstrates that TCERL outperforms state-of-the-art across a range of training subjects on four public benchmarks and is less sensitive to subject variability. The performance gain is highlighted when limited subjects are available, suggesting the robustness and transferability of the proposed method. Source code can be found in: https://github.com/haoweilou/TCERL .

Improving HPLC post-column derivatisation assays via In-Column Derivatisation

In column derivatisation (ICD) has been applied to antioxidants in food and agricultural matrices. ICD’s ability to reduce the post-column reaction loop volume is critical to investigate chromatographically before future applications are established. Due to the large number of post-column derivatisation (PCD) assays, this is out of the scope of this study to cover in one article. Hence, in this communication, we strictly focused on implementing ICD and its ability to reduce the post-column volume associated to the PCD’s reaction loop. Specifically, reversed phase (RP) separations where we take three different reaction schemes: (i) amino acids derivatised with ninhydrin, (ii) amino acids derivatised with o-phthalaldehyde (OPA), and (iii) sulfonamides derivatised with fluorescamine to exemplify three different case scenarios. This communication can be used as a practical guide for analysts. In the order of the workflow’s largest reduction in the reaction loop volume: (i) the amino acids derivatised with OPA, the ICD enabled the post-column volume to be reduced significantly from 500 µL down to 80 µL. This drastic reduction resulted in the chromatographic gain in performance – number of theoretical plates (N) (55–167 %) and peak height sensitivity (35–181 %). For the sulfonamides derivatised with fluorescamine, the ICD setup reduced the post-column volume from 500 µL down to 200 µL, and demonstrated significant improvements in efficiency – N (33–324 %), and peak height sensitivity (4–33 %). The amino acids derivatised with ninhydrin, the PCD reaction loop could not be decreased, and ICD demonstrated decreased signal to noise (S/N).

Source/drain extension asymmetric counter-doping for suppressing channel leakage in stacked nanosheet transistors

In the relentless pursuit of semiconductor device scaling, stacked silicon nanosheet gate-all-around field-effect transistors (NSFETs) are emerging as key candidates for sub-3nm technology nodes. However, the challenge of channel leakage in these devices is critical and necessitates innovative solutions. A novel SDE asymmetric counter-doping technique is proposed in this study. It investigates the impact of source/drain extension on device performance using different process schemes through three-dimensional technical computer-aided design (3D TCAD) simulations. The simulations demonstrate a comprehensive technically advantages for 25.2 %/16.65 % reduction in the off-state leakage, 27.36 %/15.03 % improvement in the on-off current ratio of N/P NSFETs, respectively. Furthermore, it shows more performance gain as the gate length scaling beyond 3 nm technology nodes. The compatibility of the asymmetric counter-doping method with mainstream NSFET integration flows and its scalability to 10 nm gate lengths indicate that it is a promising approach to optimize NSFETs performance with little extra process cost.