没有合适的资源?快使用搜索试试~ 我知道了~
资源推荐
资源详情
资源评论
Rapid Object Detection using a Boosted Cascade of Simple Features
(PAMI 2001)
Robust Real-Time Face Detection
(IJCV 2004)
by
Paul Viola & Michael Jones
•
Describes a machine learning approach for visual
object detection and construct a framework for robust
and extremely rapid object detection.
The goal of the papers
•
Integral Image;
•
Learning algorithm based on AdaBoost which is used
to select a small set of features and train the
classifiers;
•
Construct Cascade structure which dramatically
increase the speed of the detector.
Three Key Contributions
•
Let’s talk about the features first…
•
What kind of features are used in the papers?
•
What do these features stand for?
•
How do these features work?
•
What do we want from these features?
Before three key contributions
•
Representation of the Haar-Like feature:
•
two-rectangle features: edge feature
•
three-rectangle features: line feature
•
four-rectangle features: diagonal line feature
•
How do these features work?
•
All faces share some similar properties
–
The eyes region is darker than
the cheek;
–
The nose bridge region is brighter
than the eyes.
Before three key contributions
剩余31页未读,继续阅读
资源评论
- 1dave232017-09-29不错不错不错
ZYTTAE
- 粉丝: 54
- 资源: 8
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功