中文摘要：

针对无监督的主题模型无法对图像主题进行类别标记、有监督主题模型中类别信息的标记繁琐且受主观因素影响的问题,提出了一种半监督主题模型。提取图像中与位置无关的局部特征,用尺度不变特征变换对特征进行描述,用词袋模型将人脸图像表示成一组视觉单词的集合;在基于隐含狄利克雷分配（latent Dirichlet allocation,LDA）方法中的主题-单词层分布上引入少量的类别标记指导未标记样本的分类的基础上提出半监督隐含狄利克雷分配方法。在多姿态人脸判别任务上的测试结果表明该算法比无监督LDA算法分类率高9.0%~24.7%;对于部分遮挡人脸图像、未对齐的人脸图像的分类率比多姿态主成分分析法分别提高8.8%和21.5%~39.8%。结果表明该方法在少量样本标记的情况下,性能逼近有监督的隐含狄利克雷分配方法,且适用于其它图像分类问题。

英文摘要：

Topics cannot be labeled in the unsupervised topic model,while the labeling work in supervised topic models is tedious and subjective.To solve these problems,a semi-supervised topic model was proposed.First,the location-irrelevant local features were detected and described by the scale-invariant feature transform（SIFT）,based on which images were represented by a bag of visual words.Then partial labels were introduced to the topic-word level distribution in the latent Dirichlet allocation（LDA） model to guide the classification of the unlabeled data,which resulted in a semi-supervised LDA（SSLDA） model.The validation on head pose estimation showed the classification rate of the proposed method was 9.0%~24.7% higher than that of LDA.And the pose classification rate on partially occluded and misaligned face images was 8.8% and 21.5%~39.8% higher than multi-pose PCA method.With a small amount of labeled images,the proposed SSLDA model approaches the fully supervised LDA method.And it is applicable to other image classification problems.