clifftree1的博客

做机器人以及SLAM有两本圣经，一本是大名鼎鼎的《Multiple View Geometry in Computer Vision》，另一本就是至今虽然尚未出版，但是已经在SLAM界广为流传的《State Estimation for Robotics》，这本书深入讲解了李代数的理论，以及从滤波器的角度来深入分析了机器人的状态估计方法。

非常经典的一本书This is the first textbook on pattern recognition to presentthe Bayesian viewpoint. The book presents approximate inferencealgorithms that permit fast approximate answers in situations whereexact answers are not feasible. It uses graphical models todescribe probability distributions when no other books applygraphical models to machine learning. No previous knowledge ofpattern recognition or machine learning concepts is assumed.Familiarity with multivariate calculus and basic linear algebra isrequired, and some experience in the use of probabilities would behelpful though not essential as the book includes a self-containedintroduction to basic probability theory.

神经网络和统计学习（Neural networks and statistical learning） by K.-L. Du and M.N.s. Swamy

Provides examples of applications of object detection and recognition, along with the working examples of code for solutions to real computer vision problems "Object Detection and Recognition in Digital Images" addresses key problems of CV focusing on the significant issues of object detection, tracking, and recognition in images, which are not easily found in other CV books. Chapters included are dedicated to tensor methods in computer vision, classification methods and algorithms, detection and tracking, and recognition. Throughout the book content is balanced between theory, implementation and case studies in order to provide a complete and accessible treatment of the topic. Explanations of the main theoretical ideas behind each method are provided for a thorough understanding which are augmented with a rigorous mathematical derivation of the formulas, and demonstrated working in real applications. Featured case studies focus on practical applications of the discussed methods including problems of skin and face detection, pedestrian detection, road signs recognition, hand gesture recognition, blind people navigation, and traffic surveillance.Discusses the most significant and up-to-date methods of object detection and recognition.Offers a balanced treatment between theory, implementation (mainly C++ of crucial algorithms, with some algorithms in MATLAB) and case studies detailing practical applications of the discussed methods.Companion website hosting software, example data files, test images, colour images, as well as additional material such as software manual, and solutions to exercises.