A Bayesian model for efficient visual search and recognition

被引：80
|
作者：
Elazary, Lior [1 ]
Itti, Laurent [1 ,2 ]
机构：
[1] Univ So Calif, Dept Comp Sci, Los Angeles, CA 90089 USA
[2] Univ So Calif, Grad Program Neurosci, Los Angeles, CA 90089 USA
来源：
VISION RESEARCH | 2010年 / 50卷 / 14期
基金：
美国国家科学基金会;
关键词：
Recognition; Search; Attention; Feature; Scene analysis; OBJECT RECOGNITION; EYE-MOVEMENTS; ATTENTION; SELECTION; LUMINANCE; FEATURES; COLOR;
D O I：
10.1016/j.visres.2010.01.002
中图分类号：
Q189 [神经科学];
学科分类号：
071006 ;
摘要：
Humans employ interacting bottom-up and top-down processes to significantly speed up search and recognition of particular targets. We describe a new model of attention guidance for efficient and scalable first-stage search and recognition with many objects (117,174 images of 1147 objects were tested, and 40 satellite images). Performance for recognition is on par or better than SIFT and HMAX, while being, respectively, 1500 and 279 times faster. The model is also used for top-down guided search, finding a desired object in a 5 x 5 search array within four attempts, and improving performance for finding houses in satellite images. (C) 2010 Elsevier Ltd. All rights reserved.
引用
收藏
页码：1338 / 1352
页数：15
相关论文
共 50 条
12345