Application of FPGA to real-time machine learning : hardware reservoir computers and software image processing /: hardware reservoir computers and software image processing. (2018)

• Record Type: Book
• Further Information: Note: Piotr Antonik.
• Authors: Antonik, Piotr
• Contents: Intro; Supervisor's Foreword; Abstract; Preface; Acknowledgements; Contents; 1 Introduction; 1.1 From Machine Learning to Reservoir Computing; 1.1.1 Machine Learning Algorithms; 1.1.2 Artificial Neural Networks; 1.1.3 Reservoir Computing; 1.1.4 Benchmark Tasks; 1.2 Hardware Implementations: Opto-Electronic Delay Systems; 1.2.1 Time-Multiplexing; 1.2.2 Conceptual Setup; 1.2.3 Desynchronisation; 1.2.4 Experimental Setup; 1.3 Field-Programmable Gate Arrays; 1.3.1 History; 1.3.2 Market and Applications; 1.3.3 Xilinx Virtex 6: Architecture and Operation; 1.3.4 Design Flow and Implementation Tools. 2.6.4 Equalisation of a Switching Channel2.6.5 Influence of Channel Model Parameters on Equaliser Performance; 2.7 Challenges and Solutions; 2.8 Conclusion; References; 3 Backpropagation with Photonics; 3.1 Introduction; 3.2 Backpropagation Through Time; 3.2.1 General Idea and New Notations; 3.2.2 Setting Up the Problem; 3.2.3 Output Mask Gradient; 3.2.4 Input Mask Gradient; 3.2.5 Multiple Inputs/Outputs; 3.3 Experimental Setup; 3.3.1 Online Multiplication Using Cascaded MZMs; 3.3.2 Mask Parametrisation; 3.4 FPGA Design; 3.5 Results; 3.5.1 Tasks; 3.5.2 NARMA10 and VARDEL5; 3.5.3 TIMIT. 3.5.4 Gradient Descent3.5.5 Robustness; 3.6 Challenges and Solutions; 3.7 Conclusion; References; 4 Photonic Reservoir Computer with Output Feedback; 4.1 Introduction; 4.2 Reservoir Computing with Output Feedback; 4.3 Time Series Generation Tasks; 4.3.1 Frequency Generation; 4.3.2 Random Pattern Generation; 4.3.3 Mackey-Glass Chaotic Series Prediction; 4.3.4 Lorenz Chaotic Series Prediction; 4.4 Experimental Setup; 4.5 FPGA Design; 4.6 Numerical Simulations; 4.7 Results; 4.7.1 Noisy Reservoir; 4.7.2 Frequency Generation; 4.7.3 Random Pattern Generation; 4.7.4 Mackey-Glass Series Prediction. 4.7.5 Lorenz Series Prediction4.8 Challenges and Solutions; 4.9 Conclusion; References; 5 Towards Online-Trained Analogue Readout Layer; 5.1 Introduction; 5.2 Methods; 5.3 Proposed Experimental Setup; 5.3.1 Analogue Readout Layer; 5.3.2 FPGA Board; 5.4 Numerical Simulations; 5.5 Results; 5.5.1 Linear Readout: RC Circuit; 5.5.2 Nonlinear Readout; 5.6 Conclusion; References; 6 Real-Time Automated Tissue Characterisation for Intravascular OCT Scans; 6.1 Introduction; 6.2 Feature Extraction; 6.2.1 GLCM Features; 6.2.2 Methods; 6.2.3 Operation Principle; 6.2.4 FPGA Design; 6.2.5 Results.
• Publisher Details: Cham, Switzerland : Springer
• Publication Date: 2018
• Extent: 1 online resource (xxii, 171 pages), illustrations (some color)
• Subjects: 006.3/1; Physics; Machine learning; Field programmable gate arrays; COMPUTERS -- General; Field programmable gate arrays; Machine learning; Computers -- Computer Graphics; Computers -- Intelligence (AI) & Semantics; Image processing; Artificial intelligence; Computer vision; Engineering; Artificial intelligence; Technology & Engineering -- Lasers & Photonics; Laser technology & holography; Electronic books
• Languages: English
• ISBNs: 9783319910536; 3319910531
• Related ISBNs: 9783319910529; 3319910523
• Notes: Note: Includes bibliographical references.; Note: Online resource; title from PDF title page (SpringerLink, viewed May 21, 2018).
• Access Rights: Legal Deposit; Only available on premises controlled by the deposit library and to one user at any one time; The Legal Deposit Libraries (Non-Print Works) Regulations (UK).
• Access Usage: Restricted: Printing from this resource is governed by The Legal Deposit Libraries (Non-Print Works) Regulations (UK) and UK copyright law currently in force.