(18.206.177.17) 您好!臺灣時間:2021/04/11 03:15
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:柯嘉雄
研究生(外文):Ke, Jia Xiong
論文名稱:小腦式類神經網路之硬體設計及其在色彩重現系統中之應用
論文名稱(外文):Hardware design of the CMAC-based neural network and its application on color reproduction systems
指導教授:劉濱達郭耀煌郭耀煌引用關係
指導教授(外文):Liu, Bin DaGuo, Yao Huang
學位類別:博士
校院名稱:國立成功大學
系所名稱:電機工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:1996
畢業學年度:84
語文別:中文
相關次數:
  • 被引用被引用:0
  • 點閱點閱:65
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
致謝
目錄
Figure Listing
Table Listing
中文摘要
ABSTRACT
Chapter 1
1.1 Artificial Neural Networks
1.2 Local Generalization of the CMAC-based Models
1.3 Color Reproduction
1.4 Research Motivations
1.5 Organization of Dissertation
Chapter 2 DESIGN AND IMPLEMENTATION OF STANDARD CMAC
2.1 Standard CMAC
2.2 Weight Address Mapping of CMAC
2.3 Direct Weight Address Mapping
2.4 Architecture Design for Direct Address Mapping CMAC
2.4.1 Optimizing the SDUR Operation
2.4.2 Unifying and Optimizing the C*X+Y Operation
2.4.3 Alternative Architectures for Implementing Weight Address Generator
2.5 VHDL Modeling of CMAC Chip for Color Calibration
2.5.1 Implementing the SDUR Operation
2.5.2 Mapping the Pipeline Mechanism
2.6 Synthesis Results of CMAC Chip for Color Calibration
Chapter 3 DESIGN AND IMPLEMENTATION OF HIGHER-ORDER CMAC (BCMAC)
3.1 Higher-Order CMAC --B-spline CMAC (BCMAC)
3.2 Fast B-spline Computation
3.3 Extended Direct Weight Cell Address Generation
3.4 Systolic Architecture for Addressing Weight Cells
3.4.1 Architecture Mapping
3.4.2 Configuration of a PE
3.5 Simulation Results
3.5.1 Effect of the Order of B-spline Receptive Fields
3.5.2 Approximating with Various Addressing Schemes
3.5.3 Approximating with Incomplete Training Sets
3.5.4 Effect ofQuantizationLevels
Chapter 4 HARDWARE IMPLEMENTATION OF BCMAC WITH ON-CHIP LEARNING FEATURE
4.1 Design Flow
4.2 Block Diagram of the BCMAC
4.3 Systolic Weight Cell Addressing Mechanism
4.4 Output Generation Module
4.5 Access Control Module
4.5.1 Kernel Control Block
4.5.2 Quantization Block
4.5.3 Register Access
4.5.4 Weight Access Block
4.6 Design of On-Chip Learning Module
4.6.1 Interface of Learning Module
4.6.2 Architecture of Learning Module
4.7 Synthesis Results and the Demo Board
Chapter 5 PRECISION IMPROVEMENT OF CMAC
5.1 Fuzzy CMAC (FCMAC) Model
5.1.1 Definition of FCMAC
5.1.2 Fuzzy Receptive Field Functions
5.1.3 Fuzzy Inference and Output Generation
5.1.4 LearningMethod
5.1.5 SimulationResults
5.2 Sampling Algorithms for Uniformly Distributed Training Patterns
5.2.1 Clustering Phase
5.2.2 Interpolation Phase
5.3 Adaptive Input Space Quantization
5.3.1 Desired Output Dominated Clustering Quantization
5.3.2 Hardware Implementation of Adaptive Quantization
5.4 Perturbative Learning Algorithm
Chapter 6 COLOR CORRECTION AND EXPERIMENTAL RESULTS
6.1 Fundamentals of Color
6.1.1 Additive Color System (RGB additive model)
6.1.2 Subtractive Color System (CMYor CMYK additive model)
6.1.3 ClEChromaticity Diagram
6.2 Color Correction for the Color Reproduction Environment
6.2.1 Color Errors from Scanner and Printer
6.2.2 Gamut Mismatching Among Different Color Devices
6.3 Generalized Inverse Plant Control
6.4 Experimental Results
6.4.1 BCMAC Color Correction Experiments
6.4.2 Effects of Adaptive Quantization
Chapter 7 CONCLUSIONS
References
Vita
其他
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔