Object detection, a task, in the field of computer vision faces obstacles when dealing with weather conditions such as fog, rain, snow, and low light situations. This paper provides an overview of advancements in the ...
详细信息
Machine Learning (ML) models, particularly Deep Learning (DL), have made rapid progress and achieved significant milestones across various applications, including numerous safety-critical contexts. However, these mode...
详细信息
Cyber-Physical System (CPS) devices are increasing exponentially. Lacking confidentiality creates a vulnerable network. Thus, demanding the overall system with the latest and robust solutions for the defence mechanism...
详细信息
Cyber-Physical System (CPS) devices are increasing exponentially. Lacking confidentiality creates a vulnerable network. Thus, demanding the overall system with the latest and robust solutions for the defence mechanisms with low computation cost, increased integrity, and surveillance. The proposal of a mechanism that utilizes the features of authenticity measures using the Destination Sequence Distance Vector (DSDV) routing protocol which applies to the multi-WSN (Wireless Sensor Network) of IoT devices in CPS which is developed for the Device-to-Device (D2D) authentication developed from the local-chain and public chain respectively combined with the Software Defined Networking (SDN) control and monitoring system using switches and controllers that will route the packets through the network, identify any false nodes, take preventive measures against them and preventing them for any future problems. Next, the system is powered by Blockchain cryptographic features by utilizing the TrustChain features to create a private, secure, and temper-free ledger of the transactions performed inside the network. Results are achieved in the legitimate devices connecting to the network, transferring their packets to their destination under supervision, reporting whenever a false node is causing hurdles, and recording the transactions for temper-proof records. Evaluation results based on 1000+ transactions illustrate that the proposed mechanism not only outshines most aspects of Cyber-Physical systems but also consumes less computation power with a low latency of 0.1 seconds only.
The proposed work delves into the comprehensive design, development, and evaluation process of a 3D printed prosthetic arm, employing servo motors controlled by an Arduino Uno microcontroller to enable gesture mimicry...
详细信息
In this paper, we introduce the models developed by the NICA group for the sEXism Identification in Social neTworks (EXIST) Shared Task at CLEF 2024. Our participation spanned across five tasks: Sexism Identification ...
详细信息
This research investigates the relationship between music and human emotions, focusing on emotion classification in Turkish music and the role of acoustics in music emotion recognition (MER). Leveraging a unique datas...
详细信息
This research aims to develop a new approach to increase the safety and reliability of Autonomous Vehicle (AV) through the proposed risk assessment framework, supported by the trust evaluation approach derived from a ...
详细信息
Modern recommendation systems are widely used in modern data *** random and sparse embedding lookup operations are the main performance bottleneck for processing recommendation systems on traditional platforms as they...
详细信息
Modern recommendation systems are widely used in modern data *** random and sparse embedding lookup operations are the main performance bottleneck for processing recommendation systems on traditional platforms as they induce abundant data movements between computing units and ***-based processing-in-memory(PIM)can resolve this problem by processing embedding vectors where they are ***,the embedding table can easily exceed the capacity limit of a monolithic ReRAM-based PIM chip,which induces off-chip accesses that may offset the PIM ***,we deploy the decomposed model on-chip and leverage the high computing efficiency of ReRAM to compensate for the decompression performance *** this paper,we propose ARCHER,a ReRAM-based PIM architecture that implements fully yon-chip recommendations under resource ***,we make a full analysis of the computation pattern and access pattern on the decomposed *** on the computation pattern,we unify the operations of each layer of the decomposed model in multiply-and-accumulate *** on the access observation,we propose a hierarchical mapping schema and a specialized hardware design to maximize resource *** the unified computation and mapping strategy,we can coordinatethe inter-processing elements *** evaluation shows that ARCHER outperforms the state-of-the-art GPU-based DLRM system,the state-of-the-art near-memory processing recommendation system RecNMP,and the ReRAM-based recommendation accelerator REREC by 15.79×,2.21×,and 1.21× in terms of performance and 56.06×,6.45×,and 1.71× in terms of energy savings,respectively.
To meet the challenge of camouflaged object detection （COD）,which has a high degree of intrinsic similarity between the object and background,this paper proposes a double-branch fusion network（DBFN）with a parallel...
详细信息
To meet the challenge of camouflaged object detection （COD）,which has a high degree of intrinsic similarity between the object and background,this paper proposes a double-branch fusion network（DBFN）with a parallel attention selection mechanism （PASM）.In detail,a schismatic receptive field block（SRF）combined with an attention mechanism for low-level information is performed to learn texture features in one branch,and an integration of the SRF,a hybrid attention mechanism （HAM）,and a depth feature polymerization module （DFPM）is employed for high-level information to extract detection features in the other ***,both texture features and detection features are input into the PASM to acquire selective expression ***,the final result is obtained after further selective matrix optimization with atrous spatial pyramid pooling （ASPP）and a residual channel attention block （RCAB）being applied *** results on three public datasets verify that our method outperforms the state-of-the-art methods in terms of four evaluation metrics,i.e.,mean absolute error （MAE）,weighted F βmeasure （Fβω）,structural measure （Sα）,and E-measure （Eφ）
Encryption of a plaintext involves a secret key. The secret key of classical cryptosystems can be successfully determined by utilizing metaheuristic techniques. Monoalphabetic cryptosystem is one of the famous classic...
详细信息
暂无评论