Advanced search

Showing result 1 - 5 of 22 essays matching the above criteria.

  1. 1. EMONAS : Evolutionary Multi-objective Neuron Architecture Search of Deep Neural Network

    University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Author : Jiayi Feng; [2023]
    Keywords : DNN Deep Neural Network ; NAS Neural Architecture Search ; EA Evolutionary Algorithm ; Multi-Objective Optimization; Binary One Optimization; Embedded Systems; DNN Deep Neural Network ; NAS Neural Architecture Search ; EA Evolutionary Algorithm ; Multi-Objective Optimization; Binary One Optimization; Inbyggda system;

    Abstract : Customized Deep Neural Network (DNN) accelerators have been increasingly popular in various applications, from autonomous driving and natural language processing to healthcare and finance, etc. However, deploying them directly on embedded system peripherals within real-time operating systems (RTOS) is not easy due to the paradox of the complexity of DNNs and the simplicity of embedded system devices. READ MORE

  3. 2. Leveraging Posits for the Conjugate Gradient Linear Solver on an Application-Level RISC-V Core

    University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Author : David Mallasén Quintana; [2022]
    Keywords : Computer Arithmetic; Conjugate Gradient; Posit; IEEE-754; Floating Point; High-Performance Computing; RISC-V; Datoraritmetik; Konjugerad Gradient; Posit; IEEE-754; Flytande Punkt; Högpresterande Beräkningar; RISC-V;

    Abstract : Emerging floating-point arithmetics provide a way to optimize the execution of computationally-intensive algorithms. This is the case with scientific computational kernels such as the Conjugate Gradient (CG) linear solver. Exploring new arithmetics is of paramount importance to maximize the accuracy and timing performance of these algorithms. READ MORE

  4. 3. Representation and Efficient Computation of Sparse Matrix for Neural Networks in Customized Hardware

    University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Author : Lihao Yan; [2022]
    Keywords : Convolutional neural networks; Sparse matrix representation; Model compression; Algorithm-hardware co-design; AlexNet; Konvolutionella neurala nätverk; Sparsam matrisrepresentation; Modellkomprimering; Algoritm-hårdvarusamdesign; AlexNet;

    Abstract : Deep Neural Networks are widely applied to various kinds of fields nowadays. However, hundreds of thousands of neurons in each layer result in intensive memory storage requirement and a massive number of operations, making it difficult to employ deep neural networks on mobile devices where the hardware resources are limited. READ MORE

  5. 4. Mitchell-Based Approximate Operations on Floating-Point Numbers

    University essay from Linköpings universitet/Datorteknik

    Author : Noah Hellman; [2021]
    Keywords : floating-point numbers; approximations; arithmetics; asic; fpga; mitchell;

    Abstract : By adapting Mitchell's algorithm for floating-point numbers, one can efficiently perform arithmetic floating-point operations in an approximate logarithmic domain in order to perform approximate computations of functions such as multiplication, division, square root and others. This work examines how this algorithm can be improved in terms of accuracy and hardware complexity by applying a set of various methods that are parametrized and offer a large design space. READ MORE

  6. 5. FPGA Implementation of an Anonymization Algorithm

    University essay from Lunds universitet/Institutionen för elektro- och informationsteknik

    Author : Niklas Kjellman; Rehenuma Tarannum; [2020]
    Keywords : FPGA anonymization privacy anonymization filter spacial filter; Technology and Engineering;

    Abstract : With the large amount of surveillance cameras in our public spaces there has been much discussion about their effects on privacy. Axis communications has an algorithm that tries to remedy this by making it hard to see who is in an image but still have it possible for a neural network to detect that there is a person present. READ MORE