Feature Extraction and Image Processing for Computer Vision 4/e (美國原版)

Nixon, Mark, Aguado, Alberto S.

  • 出版商: Academic Press
  • 出版日期: 2019-11-18
  • 售價: $3,090
  • 貴賓價: 9.5$2,936
  • 語言: 英文
  • 頁數: 632
  • 裝訂: Quality Paper - also called trade paper
  • ISBN: 0128149760
  • ISBN-13: 9780128149768
  • 相關分類: Computer Vision
  • 海外代購書籍(需單獨結帳)

買這商品的人也買了...

商品描述

Feature Extraction and Image Processing for Computer Vision, Fourth Edition, describes basic principles, theories and techniques used to design visual search and recognition engines. The book explores how visual features are extracted and quantized, how an indexing system is built, includes comparisons to alternative schemes, and presents cutting-edge techniques, like deep learning. From the feature end, users will find out how hash functions transfer high-dimensional feature descriptors into compact binary code. From the model end, the book discusses recent advances in compressing and accelerating large, deep learning models (like convolutional neural networks) to fit into mobile and robotic memory storage.

 

  • Introduces the building blocks of visual search and recognition engines
  • Provides detailed reviews of visual feature hashing
  • Contains reviews of deep learning model compression
  • Presents the latest cutting-edge research, such as compact descriptors for visual search and visual feature hashing

商品描述(中文翻譯)

《Feature Extraction and Image Processing for Computer Vision, Fourth Edition》描述了設計視覺搜索和識別引擎所使用的基本原理、理論和技術。本書探討了視覺特徵的提取和量化方法,建立索引系統的過程,並與其他方案進行了比較,還介紹了深度學習等尖端技術。從特徵的角度來看,讀者將了解哈希函數如何將高維特徵描述符轉換為緊湊的二進制代碼。從模型的角度來看,本書討論了壓縮和加速大型深度學習模型(如卷積神經網絡)以適應移動和機器人記憶存儲的最新進展。

本書的主要內容包括:
- 介紹視覺搜索和識別引擎的基本組件
- 詳細評論視覺特徵哈希
- 回顧深度學習模型壓縮的研究
- 提供最新的尖端研究成果,如視覺搜索的緊湊描述符和視覺特徵哈希