The convolutional neural network (CNN) is widely utilized in computer vision due to its ability to effectively harness correlation information within data. However, when the data's dimensionality or the model'...
详细信息
ISBN:
(数字)9798350364637
ISBN:
(纸本)9798350364644
The convolutional neural network (CNN) is widely utilized in computer vision due to its ability to effectively harness correlation information within data. However, when the data's dimensionality or the model's complexity becomes excessively large, CNN faces significant challenges in maintaining efficient learning. The quantum convolutional neural network (QCNN) offers an innovative solution by leveraging quantum computing environments, to address problems conventionally solved by CNN or enhance the performance of existing learning models. This paper proposes a model of the CNN structure for quantum computing and introduces a method for enhancing model performance by integrating heterogeneous knowledge transfer (HKT) for QCNN with the CNN model used in conventional computer vision tasks. Furthermore, this paper examines the feasibility of the QCNN-HKT model in comparison to CNN and QCNN by conducting training experiments on the KITTI dataset with quantum computing simulation libraries Torchquantum.
As the computing capability of mobile devices increases and the web technologies are more developed, smart phones have an almost similar performance to desktop computers of the past and allow user to search what he wa...
As the computing capability of mobile devices increases and the web technologies are more developed, smart phones have an almost similar performance to desktop computers of the past and allow user to search what he want to find without any difficulty in embedded web browser. In addition, multi-touch interface equipped with them is very intuitive and immediate. This paper presents a novel remote control method using mobile phone to h a ndl e w eb contents and change their positions in a large display screen such as smart TV, digital signage screen etc.
The role of Numerical Integration in the evaluation of definite improper integrals is being increasingly appreciated as there are no simple analytical results available. In this paper the authors explore four such qua...
详细信息
The role of Numerical Integration in the evaluation of definite improper integrals is being increasingly appreciated as there are no simple analytical results available. In this paper the authors explore four such quadrature formulae and their performance in evaluating Logarithmic integrals, a class of definite improper integrals and one of the important integrals in Number Theory. The performance of the proposed methods are compared with some well known quadrature formulae like Simpson's rule, Trapezoidal rule , Weddle's rule etc.
This paper presents a quick survey of Synchronous and Asynchronous waterfilling algorithm for Multi-user MIMO interference systems. Presenting the game theoretic formulation for non-cooperative interference multi-user...
This paper presents a quick survey of Synchronous and Asynchronous waterfilling algorithm for Multi-user MIMO interference systems. Presenting the game theoretic formulation for non-cooperative interference multi-user MIMO channel, comparison and analysis of the main iterative algorithms; Simultaneous, Sequential, Asynchronous and Totally Asynchronous iterative waterfilling distributed algorithms (IWFA). Simulation results reveal the characteristics speed of convergence of the algorithms.
With chaotic sequence optimization algorithm,the function optimization problem for analyzing data pumping test conducted in aquifer to estimate such aquifer parameters as transmissivity and storage co
With chaotic sequence optimization algorithm,the function optimization problem for analyzing data pumping test conducted in aquifer to estimate such aquifer parameters as transmissivity and storage co
Theoretical and some practical problems of evolution strategies (ES) construction are considered. In contrast to genetic algorithms ES reflect the evolution progress on the phenotype level. The ES basic scheme is anal...
详细信息
Theoretical and some practical problems of evolution strategies (ES) construction are considered. In contrast to genetic algorithms ES reflect the evolution progress on the phenotype level. The ES basic scheme is analyzed: variants of basic algorithm modification and adaptation for different optimization tasks are investigated. The problems of evaluation of selection schemes and ES convergencecomputing complexity are analyzed.
Classic McEliece is a Code-based Key Encapsulation Mechanisms (KEM) and one of the candidate algorithms in the NIST PQC competition. Based on the McEliece cryptosystem developed in 1978, this system relies on the Nied...
详细信息
ISBN:
(纸本)9789819980239;9789819980246
Classic McEliece is a Code-based Key Encapsulation Mechanisms (KEM) and one of the candidate algorithms in the NIST PQC competition. Based on the McEliece cryptosystem developed in 1978, this system relies on the Niederreiter variant of McEliece. It consists of three phases: Key Generation, Encapsulation, and Decapsulation. In this paper, we propose an optimized implementation of the internal multiplication operations of Classic McEliece on the ARMv8 processor. We utilize parallel computing techniques using vector registers and vector instructions of the ARMv8 processor. We specifically focus on optimizing the multiplication operation, which is a major contributor to the overall execution time of the Classic McEliece algorithm, by leveraging the commutative property and implementing an parallelization technique. As a result, our approach achieves a maximum performance improvement of 2.82x compared to the reference implementation in the multiplication operation.
USN and RFID, as new technologies for the realization of ubiquitous computing, are technically closely related with each other. In this paper, we propose an efficient processing method of various kinds of sensor data ...
详细信息
USN and RFID, as new technologies for the realization of ubiquitous computing, are technically closely related with each other. In this paper, we propose an efficient processing method of various kinds of sensor data used in USN environment to provide efficiency from the view point of not only the convergence of USN and RFID technologies but also on cost and extensibility. Specifically, we propose a method of converting various sensor data into EPC data format which can be handled in ALE-compliant RFID middleware. By using the proposed method, existing ALE-compliant middleware based on international standard specification can easily handle not only RFID tag data but also general sensor data simultaneously without any changes in the middleware. And in addition, various types of ubiquitous application services can be implemented at a low cost and in a shorter time.
This paper presents a branch prediction algorithm and a 4-way set-associative cache for performance improvement of 32-bit RISC processor and a clock gating algorithm using ODC (observability don't care) operation ...
详细信息
This paper presents a branch prediction algorithm and a 4-way set-associative cache for performance improvement of 32-bit RISC processor and a clock gating algorithm using ODC (observability don't care) operation for a low-power processor. The branch prediction algorithm has a structure using BTB (branch target buffer) and 4-way set associative cache using pseudo LRU (least recently used) algorithm. The proposed algorithm is applied to OpenRISC1200 processor, embedded processor and implemented on Xilinx VIRTEX-4 XC4VLX80 FPGA device and the FPGA executes at the maximum frequency of 53.042MHz. As a result of estimation of performance and dynamic power, the performance of the OpenRISC1200 processor using the proposed algorithm is improved about 5~9% and dynamic power of the processor using Samsung 0.18 mum technology library is reduced by 13.9%.
The continuous physical-mathematical model of the 3D structures of protein has been obtained by analyzed the 3D structures of biological protein according to the distance and dihedral angle distribution of amino acid ...
详细信息
The continuous physical-mathematical model of the 3D structures of protein has been obtained by analyzed the 3D structures of biological protein according to the distance and dihedral angle distribution of amino acid residue conforming to Boltzmann theorem. On this foundation, stretches of the B (D) main chain for the bovine insulin have been calculated by application of the improved simulated annealing algorithm. The consistency of stretches and space folded structure for protein has been proved by edited protein sequences 1RPB folded prediction program. The global lowest energy has been calculated by amino acid sequence and a better convergence effect has been achieved.
暂无评论