Titlebook: Visual Object Recognition; Kristen Grauman,Bastian Leibe Book 2011 Springer Nature Switzerland AG 2011

meritorious

積習(xí)難改

外星人

FUME

無聊點(diǎn)好

Other Considerations and Current Challenges,ey representations, matching, learning, and detection strategies to identify visual objects. However, it is certainly not an exhaustive overview of all work ongoing in the research community. In this chapter, we point out other accompanying threads of research on issues central to recognition that a

harpsichord

Conclusions,e matching and learned generic category models. Using the basic frameworks outlined above, it is now possible to perform reliable recognition for many types of specific objects and for a number of generic object categories. As indicated by the ongoing threads discussed in the previous chapter, chall

Endemic

Book 2011ility to identify and localize categories, places, and objects. This tutorial overviews computer vision algorithms for visual object recognition and image classification. We introduce primary representations and learning approaches, with an emphasis on recent advances in the field. The target audien

避開

Matching Local Features, we essentially want to search among all previously seen local descriptors, and retrieve those that are nearest according to Euclidean distance in the feature space (such as the 128-dimensional “SIFT space”).

按時(shí)間順序

Representations for Object Categories,e descriptors for the appearance of a set of local parts together with a geometric layout model (Section 8.2). In terms of appearance, there is significant overlap in the low-level feature used for either type of model; however, we will see in Chapter 9 that their associated procedures for detecting novel instances within a scene differ.

Fecal-Impaction