Hasil untuk "Architecture"

Yue Liu, Yunjie Tian, Yuzhong Zhao et al.

Designing computationally efficient network architectures remains an ongoing necessity in computer vision. In this paper, we adapt Mamba, a state-space language model, into VMamba, a vision backbone with linear time complexity. At the core of VMamba is a stack of Visual State-Space (VSS) blocks with the 2D Selective Scan (SS2D) module. By traversing along four scanning routes, SS2D bridges the gap between the ordered nature of 1D selective scan and the non-sequential structure of 2D vision data, which facilitates the collection of contextual information from various sources and perspectives. Based on the VSS blocks, we develop a family of VMamba architectures and accelerate them through a succession of architectural and implementation enhancements. Extensive experiments demonstrate VMamba's promising performance across diverse visual perception tasks, highlighting its superior input scaling efficiency compared to existing benchmark models. Source code is available at https://github.com/MzeroMiko/VMamba.

Rahima Khanam, Muhammad Hussain

This study presents an architectural analysis of YOLOv11, the latest iteration in the YOLO (You Only Look Once) series of object detection models. We examine the models architectural innovations, including the introduction of the C3k2 (Cross Stage Partial with kernel size 2) block, SPPF (Spatial Pyramid Pooling - Fast), and C2PSA (Convolutional block with Parallel Spatial Attention) components, which contribute in improving the models performance in several ways such as enhanced feature extraction. The paper explores YOLOv11's expanded capabilities across various computer vision tasks, including object detection, instance segmentation, pose estimation, and oriented object detection (OBB). We review the model's performance improvements in terms of mean Average Precision (mAP) and computational efficiency compared to its predecessors, with a focus on the trade-off between parameter count and accuracy. Additionally, the study discusses YOLOv11's versatility across different model sizes, from nano to extra-large, catering to diverse application needs from edge devices to high-performance computing environments. Our research provides insights into YOLOv11's position within the broader landscape of object detection and its potential impact on real-time computer vision applications.

Laith Alzubaidi, Jinglan Zhang, A. Humaidi et al.

In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

Bowen Cheng, Ishan Misra, A. Schwing et al.

Image segmentation groups pixels with different semantics, e.g., category or instance membership. Each choice of semantics defines a task. While only the semantics of each task differ, current research focuses on designing spe-cialized architectures for each task. We present Masked- attention Mask Transformer (Mask2Former), a new archi-tecture capable of addressing any image segmentation task (panoptic, instance or semantic). Its key components in-clude masked attention, which extracts localized features by constraining cross-attention within predicted mask regions. In addition to reducing the research effort by at least three times, it outperforms the best specialized architectures by a significant margin on four popular datasets. Most no-tably, Mask2Former sets a new state-of-the-art for panoptic segmentation (57.8 PQ on COCO), instance segmentation (50.1 AP on COCO) and semantic segmentation (57.7 mIoU onADE20K).

Xiangxiang Chu, Zhi Tian, Yuqing Wang et al.

Very recently, a variety of vision transformer architectures for dense prediction tasks have been proposed and they show that the design of spatial attention is critical to their success in these tasks. In this work, we revisit the design of the spatial attention and demonstrate that a carefully-devised yet simple spatial attention mechanism performs favourably against the state-of-the-art schemes. As a result, we propose two vision transformer architectures, namely, Twins-PCPVT and Twins-SVT. Our proposed architectures are highly-efficient and easy to implement, only involving matrix multiplications that are highly optimized in modern deep learning frameworks. More importantly, the proposed architectures achieve excellent performance on a wide range of visual tasks, including image level classification as well as dense detection and segmentation. The simplicity and strong performance suggest that our proposed architectures may serve as stronger backbones for many vision tasks. Our code is released at https://github.com/Meituan-AutoML/Twins .

Jeya Maria Jose Valanarasu, Poojan Oza, I. Hacihaliloglu et al.

Over the past decade, Deep Convolutional Neural Networks have been widely adopted for medical image segmentation and shown to achieve adequate performance. However, due to the inherent inductive biases present in the convolutional architectures, they lack understanding of long-range dependencies in the image. Recently proposed Transformer-based architectures that leverage self-attention mechanism encode long-range dependencies and learn representations that are highly expressive. This motivates us to explore Transformer-based solutions and study the feasibility of using Transformer-based network architectures for medical image segmentation tasks. Majority of existing Transformer-based network architectures proposed for vision applications require large-scale datasets to train properly. However, compared to the datasets for vision applications, for medical imaging the number of data samples is relatively low, making it difficult to efficiently train transformers for medical applications. To this end, we propose a Gated Axial-Attention model which extends the existing architectures by introducing an additional control mechanism in the self-attention module. Furthermore, to train the model effectively on medical images, we propose a Local-Global training strategy (LoGo) which further improves the performance. Specifically, we operate on the whole image and patches to learn global and local features, respectively. The proposed Medical Transformer (MedT) is evaluated on three different medical image segmentation datasets and it is shown that it achieves better performance than the convolutional and other related transformer-based architectures. Code: https://github.com/jeya-maria-jose/Medical-Transformer

James C. Phillips, David J. Hardy, Julio D C Maia et al.

NAMDis a molecular dynamics program designed for high-performance simulations of very large biological objects on CPU- and GPU-based architectures. NAMD offers scalable performance on petascale parallel supercomputers consisting of hundreds of thousands of cores, as well as on inexpensive commodity clusters commonly found in academic environments. It is written in C++ and leans on Charm++ parallel objects for optimal performance on low-latency architectures. NAMD is a versatile, multipurpose code that gathers state-of-the-art algorithms to carry out simulations in apt thermodynamic ensembles, using the widely popular CHARMM, AMBER, OPLS, and GROMOS biomolecular force fields. Here, we review the main features of NAMD that allow both equilibrium and enhanced-sampling molecular dynamics simulations with numerical efficiency. We describe the underlying concepts utilized by NAMD and their implementation, most notably for handling long-range electrostatics; controlling the temperature, pressure, and pH; applying external potentials on tailored grids; leveraging massively parallel resources in multiple-copy simulations; and hybrid quantum-mechanical/molecular-mechanical descriptions. We detail the variety of options offered by NAMD for enhanced-sampling simulations aimed at determining free-energy differences of either alchemical or geometrical transformations and outline their applicability to specific problems. Last, we discuss the roadmap for the development of NAMD and our current efforts toward achieving optimal performance on GPU-based architectures, for pushing back the limitations that have prevented biologically realistic billion-atom objects to be fruitfully simulated, and for making large-scale simulations less expensive and easier to set up, run, and analyze. NAMD is distributed free of charge with its source code at www.ks.uiuc.edu.

Yong Yu, Xiaosheng Si, Changhua Hu et al.

Recurrent neural networks (RNNs) have been widely adopted in research areas concerned with sequential data, such as text, audio, and video. However, RNNs consisting of sigma cells or tanh cells are unable to learn the relevant information of input data when the input gap is large. By introducing gate functions into the cell structure, the long short-term memory (LSTM) could handle the problem of long-term dependencies well. Since its introduction, almost all the exciting results based on RNNs have been achieved by the LSTM. The LSTM has become the focus of deep learning. We review the LSTM cell and its variants to explore the learning capacity of the LSTM cell. Furthermore, the LSTM networks are divided into two broad categories: LSTM-dominated networks and integrated LSTM networks. In addition, their various applications are discussed. Finally, future research directions are presented for LSTM networks.

J. Dongarra, Mark Gates, A. Haidar et al.

The recent version of the Parallel Linear Algebra Software for Multicore Architectures (PLASMA) library is based on tasks with dependencies from the OpenMP standard. The main functionality of the library is presented. Extensive benchmarks are targeted on three recent multicore and manycore architectures, namely, an Intel Xeon, Intel Xeon Phi, and IBM POWER 8 processors.

Asifullah Khan, A. Sohail, Umme Zahoora et al.

Deep Convolutional Neural Network (CNN) is a special type of Neural Networks, which has shown exemplary performance on several competitions related to Computer Vision and Image Processing. Some of the exciting application areas of CNN include Image Classification and Segmentation, Object Detection, Video Processing, Natural Language Processing, and Speech Recognition. The powerful learning ability of deep CNN is primarily due to the use of multiple feature extraction stages that can automatically learn representations from the data. The availability of a large amount of data and improvement in the hardware technology has accelerated the research in CNNs, and recently interesting deep CNN architectures have been reported. Several inspiring ideas to bring advancements in CNNs have been explored, such as the use of different activation and loss functions, parameter optimization, regularization, and architectural innovations. However, the significant improvement in the representational capacity of the deep CNN is achieved through architectural innovations. Notably, the ideas of exploiting spatial and channel information, depth and width of architecture, and multi-path information processing have gained substantial attention. Similarly, the idea of using a block of layers as a structural unit is also gaining popularity. This survey thus focuses on the intrinsic taxonomy present in the recently reported deep CNN architectures and, consequently, classifies the recent innovations in CNN architectures into seven different categories. These seven categories are based on spatial exploitation, depth, multi-path, width, feature-map exploitation, channel boosting, and attention. Additionally, the elementary understanding of CNN components, current challenges, and applications of CNN are also provided.

Shawn Hershey, Sourish Chaudhuri, D. Ellis et al.

Convolutional Neural Networks (CNNs) have proven very effective in image classification and show promise for audio. We use various CNN architectures to classify the soundtracks of a dataset of 70M training videos (5.24 million hours) with 30,871 video-level labels. We examine fully connected Deep Neural Networks (DNNs), AlexNet [1], VGG [2], Inception [3], and ResNet [4]. We investigate varying the size of both training set and label vocabulary, finding that analogs of the CNNs used in image classification do well on our audio classification task, and larger training and label sets help up to a point. A model using embeddings from these classifiers does much better than raw features on the Audio Set [5] Acoustic Event Detection (AED) classification task.

Guillaume Lample, Miguel Ballesteros, Sandeep Subramanian et al.

Comunicacio presentada a la 2016 Conference of the North American Chapter of the Association for Computational Linguistics, celebrada a San Diego (CA, EUA) els dies 12 a 17 de juny 2016.

Bowen Baker, O. Gupta, Nikhil Naik et al.

At present, designing convolutional neural network (CNN) architectures requires both human expertise and labor. New architectures are handcrafted by careful experimentation or modified from a handful of existing networks. We introduce MetaQNN, a meta-modeling algorithm based on reinforcement learning to automatically generate high-performing CNN architectures for a given learning task. The learning agent is trained to sequentially choose CNN layers using $Q$-learning with an $\epsilon$-greedy exploration strategy and experience replay. The agent explores a large but finite space of possible architectures and iteratively discovers designs with improved performance on the learning task. On image classification benchmarks, the agent-designed networks (consisting of only standard convolution, pooling, and fully-connected layers) beat existing networks designed with the same layer types and are competitive against the state-of-the-art methods that use more complex layer types. We also outperform existing meta-modeling approaches for network design on image classification tasks.

Kateřina Bruštíková, Jitka Forstová, Barbora Holajová et al.

Abstract BK polyomavirus (BKPyV) is a human pathogen that causes severe disease in immunocompromised individuals. Although discovered in the 1970s, important gaps in our understanding of BKPyV biology persist. Key unresolved areas include the precise molecular mechanisms governing viral latency and reactivation, the specific host and viral factors determining the virus tropism towards the urinary tract, and the intricate virus-host interactions that drive clinical pathogenesis. These unresolved biological questions have stalled the development of targeted therapeutics; as a result, no specific antiviral therapy is currently available for BKPyV-related diseases. In this review, we examine findings from both experimental models and clinical samples that investigate how BKPyV remodels host organelles and the molecular pathways underlying these alterations. We focus on BKPyV-driven changes in cellular membranes, including endoplasmic reticulum remodeling, mitochondrial disruption, the formation of endoplasmic reticulum-derived tubuloreticular structures, vacuoles, and autophagosomes, as well as the accumulation of lipid droplets. Collectively, these organelle-specific modifications highlight membrane remodeling as a central feature of BKPyV replication and pathogenesis. Addressing the key knowledge gaps in the molecular basis of virus-induced membrane remodeling will be critical for guiding the development of effective antiviral strategies.

Junwei Wang, Fuqiang Zhao, Zihan Guo et al.

Composite rubber bearing is an important supporting component in bridge structure system, its aging and shear performance will affect the safety of the whole structure. However, due to the complexity of LRB specifications and sizes, the shear properties of aging LRB under different parameters were studied. In this study, the thermal aging and shear tests of 12 LRBs of the same specifications were first carried out, and the test results were taken as a reference, and the finite element model was established to select the constitutive model and determine the parameters, and finally the constitutive model and parameters consistent with the test were determined. Then, LRBs with different shape coefficient, diameter and number of layers were established, and shear simulation was carried out respectively to compare with the shear performance of the test supports, and the changes of parameters such as maximum shear force, energy dissipation, equivalent shear stiffness, initial sliding displacement and sliding distance generated by LRBs of different specifications at different shear stages were studied. The results show that for LRB of the same specifications, aging does not affect the maximum shear force, but the hardness and energy dissipation of rubber material increase with the aging time, and the initial sliding distance decreases with the aging time. For LRB with different parameters, under the same aging time, the maximum shear force and energy dissipation increase with the increase of shear deformation, and the equivalent shear stiffness decreases with the increase of shear degree. The maximum shear force, energy dissipation and initial shear stiffness of LRB increase with the increase of shape coefficient and diameter. The number of layers of the LRB does not affect the maximum shear force, but the energy dissipation increases with the increase of the number of layers, and the equivalent shear stiffness decreases with the increase of the number of layers. The larger the shape factor, diameter and layer number of LRB, the more likely it is to slip. Therefore, the influence of bearing parameters on the shear performance of LRB should be considered comprehensively when designing LRB in actual engineering.

Xuran Men, Jinlong He, Bo Jiao et al.

With the rapid development and accelerated utilization of marine resources, multi-body floating systems have become extensively used in practical applications. This study examines the coupled motions of a side-by-side anchoring system for five fishing vessels in a harbor using ANSYS-AQWA. The system is connected by hawsers and equipped with fenders to reduce collisions between the vessels. It is designed to operate in the sheltered wind-wave combined environment within Ningbo Zhoushan Port, China. Considering the diverse types and quantities of fishing vessels in the anchorage area, this paper proposes a mixed arrangement of three large-scale fishing vessels in the middle and two small-scale vessels on both sides. The time-domain analysis is performed on this system under the combined effects of wind and waves, calculating the motion responses of the five fishing vessels along with the mechanical loads at the hawsers, fenders, and moorings. The results indicate that the maximum loads on these mechanical components remain well within the safe working limits, ensuring reliable operation. In addition, the impact of varying wind-wave angles on the coupled motions of the fishing vessel system are studied. As the wind-wave angle increases, the surge motion of the fishing vessels gradually decreases, while the sway motion intensifies. The forces on the hawsers, fenders, and mooring system exhibit distinct characteristics at different angles.

Architectural Intelligence

Castillo-González, Humberto, Slot, Jason C., Yarwood, Stephanie et al.

Fungal endophytes play a pivotal role in tropical forest dynamics, influencing plant fitness through growth stimulation, disease suppression, stress tolerance, and nutrient mobilization. This study investigates the effects of region, leaf developmental stage, and tissue type on endophyte communities in tropical plants. Young and mature leaves were collected from 47 Rubiaceae species, and sapwood from 23 species, in old-growth forests of Golfito and Guanacaste, Costa Rica. Fungal diversity and composition were assessed through metabarcoding of the ITS2 nrDNA region. Most identified ASVs belonged to the phylum Ascomycota. The orders Botryosphaeriales and Glomerellales significantly contributed to endophytic assemblages, without detection of host-specific communities. We observed significant differences in species richness across regions, confirming distinct compositions through beta diversity. No statistically significant variances were found between mature and juvenile leaf tissues. In contrast, leaves exhibited richer and more diverse assemblages than sapwood. As plants experienced diverse environments over time and space, our results may be influenced by changing structural and chemical properties through ontogeny. Given the potential impact of these fungi on agricultural and forest ecosystems, ongoing research is crucial to discern the roles of hosts, endophytes, and other ecological mechanisms in apparent colonization patterns.

Zhaoxuan Lu, Lyuchao Liao, Xingang Xie et al.

In recent years, climate change and marine pollution have significantly degraded coral reefs, highlighting the urgent need for automated coral detection to monitor marine ecosystems. However, underwater coral detection presents unique challenges, including low image contrast, complex coral structures, and dense coral growth, which limit the effectiveness of general object detection algorithms. To address these challenges, we propose SCoralDet, a soft coral detection model based on the YOLO architecture. First, we introduce a Multi-Path Fusion Block (MPFB) to capture coral features across multiple scales, enhancing the model’s robustness to uneven lighting and image blurring. We further improve inference efficiency by applying reparameterization. Second, we integrate lightweight components such as GSConv and VoV-GSCSP to reduce computational overhead without sacrificing performance. Additionally, we develop an Adaptive Power Transformation label assignment strategy, which dynamically adjusts anchor alignment metrics. By incorporating soft labels and soft central region loss, our model is guided to prioritize high-quality, well-aligned predictions. We evaluate SCoralDet on the Soft-Coral dataset, achieving an inference latency of 9.52 ms and an mAP50 of 81.9. This surpasses the performance of YOLOv5 (79.9), YOLOv6 (79.4), YOLOv8 (79.5), YOLOv9 (78.3), and YOLOv10 (79.5). These results demonstrate the effectiveness and practicality of SCoralDet in underwater coral detection tasks.

Shiqi ZHAO, Xiaohong HUANG, Zhigang ZHONG

To solve the problem of lack of validation for exchanging messages in BGP, a inter-domain routing mechanism, which consisted of a reputation evaluation mechanism and a reputation-based BGP optimal routing algorithm, was proposed.The reputation evaluation mechanism used a distributed autonomous system (AS) alliance architecture, which divided node routing behavior in detail.The service domain and observation weight were used as indicators to quantify the impact of node behavior.By designing a feedback mechanism, the reputation value not only reflected the good and bad of nodes, but also reflected the node’s resistance to malicious attacks.The reputation-based BGP routing selection algorithm adds a “security” policy to the existing routing selection algorithm by filtering routes containing low-reputation nodes and selecting the best route among high reputation routes.The experimental results show that the proposed mechanism outperform most existing reputation mechanisms by avoiding routes with vulnerable nodes and restraining the propagation of illegal routes, thereby providing a more secure inter-domain routing environment.