Description: 基于支持向量机的人脸检测训练集增强算法实现。根据支持向量机(support vector machine，简称SVM)~ ，对基于边界的分类算"~(geometric approach)~
言，类别边界附近的样本通常比其他样本包含有更多的分类信息．基于这一基本思路，以人脸检测问题为例．探讨了
对给定训练样本集进行边界增强的问题，并为此而提出了一种基于支持向量机和改进的非线性精简集算法
IRS(improved reduced set)的训练集边界样本增强算法，用以扩大-91l练集并改善其样本分布．其中，所谓IRS算法是指
在精简集(reduced se0算法的核函数中嵌入一种新的距离度量一一图像欧式距离一一来改善其迭代近似性能，IRS
可以有效地生成新的、位于类别边界附近的虚拟样本以增强给定训练集．为了验证算法的有效性，采用增强的样本
集训练基于AdaBoost的人脸检测器，并在MIT+CMU正面人脸测试库上进行了测试．实验结果表明通过这种方法
能够有效地提高最终分类器的人脸检测性能．-According to support vector machines(SVMs)，for those geometric approach based classification
methods，examples close to the class boundary usually are more informative than others．Taking face detection as an
example，this paper addresses the problem of enhancing given training set and presents a nonlinear method to tackle
the problem effectively．Based on SVM and improved reduced set algorithm (IRS)，the method generates new
examples lying close to the face／non—face class boundary to enlarge the original dataset and hence improve its
sample distribution．The new IRS algorithm has greatly improved the approximation performance of the original
reduced set(RS)method by embedding a new distance metric called image Euclidean distance(IMED)into the
keme1 function．To verify the generalization capability of the proposed method，the enhanced dataset is used to train
an AdaBoost．based face detector and test it on the MIT+CMU frontal face test set．The experimental results show
that the origina Platform: |
Size: 649216 |
Author:郭事业 |
Hits:
Description: 视频跟踪是视频信息处理的一个重要研究方向。交通信息管理、公安刑事侦查过程中人的跟踪占有很大比例。基于人脸特征的视频跟踪主要研究视频信息中人脸特征的提取，人脸特征的对比算法，人的运动的分析方法，及跟踪方法，跟踪结果的表示。- criminal investigation,anti-terrorism and military installations. Consequently, to recognize person identity speedy and exactly in large-scale crowd has become an important way to protect the public social security, guarantee national harmony and reinforce pre-warning capability in public safety.This paper designs a real time automatic face detection, tracking and recognition system for video, which can detect, track faces and recognize human identity within the scope of video. The system consists of three parts, which are multi-face detection, multi-face tracking and identity recognition.
For face detection, this paper presents an face detection algorithm called Adaboost-ASM face detection algorithm, combines Adaboost face detection algorithm with the adaptive shape model (ASM), to solve the problem that Adaboost face detection algorithm recognizes the non-face region and complex region as a face region mistakenly.Finally, the Adaboost-ASM algorithm achieves the real time face Platform: |
Size: 641024 |
Author:bai |
Hits:
Search - adaboost face non face