A novel hierarchical Bag-of-Words model for compact action representation

Sun, Quansen; Liu, Hong; Ma, Liqian; Zhang, Tianwei

doi:10.1016/j.neucom.2015.09.074

Cited by 15 publications

(3 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The vector is compact and is invariant to translation, rotation, and scaling [21]. However, where the codebook size is very large, the dimension of the image vector can also become very large [112]. This will call for the inclusion of dimension reduction methods; thereby achieve better retrieval results than BOVW at a high computational cost [113].…”

Section: Image Representationsmentioning

confidence: 99%

Review of the application of machine learning to the automatic semantic annotation of images

Olaode

Naghdy

2019

IET Image Processing

View full text Add to dashboard Cite

Section: Image Representationsmentioning

confidence: 99%

Review of the application of machine learning to the automatic semantic annotation of images

Olaode

Naghdy

2019

IET Image Processing

View full text Add to dashboard Cite

“…Containing the term will be represented as '1', while the absence will be represented as '0'. Now each data instance will be represented as a vector like the following vector: 00001000000 in this regard, each vector will contains a value of '1' which refers to the occurrence of a corresponding term [20,21].…”

Section: N-gram Representationmentioning

confidence: 99%

Classifying confidential data using SVM for efficient cloud query processing

Tayyeh

AL-Jumaili

2019

TELKOMNIKA

View full text Add to dashboard Cite

Nowadays, organizations are widely using a cloud database engine from the cloud service providers. Privacy still is the main concern for these organizations where every organization is strictly looking forward more secure environment for their own data. Several studies have proposed different types of encryption methods to protect the data over the cloud. However, the daily transactions represented by queries for such databases makes encryption is inefficient solution. Therefore, recent studies presented a mechanism for classifying the data prior to migrate into the cloud. This would reduce the need of encryption which enhances the efficiency. Yet, most of the classification methods used in the literature were based on string-based matching approach. Such approach suffers of the exact match of terms where the partial matching would not be considered. This paper aims to take the advantage of N-gram representation along with Support Vector Machine classification. A real-time data will used in the experiment. After conducting the classification, the Advanced Encryption Standard algorithm will be used to encrypt the confidential data. Results showed that the proposed method outperformed the baseline encryption method. This emphasizes the usefulness of using the machine learning techniques for the process of classifying the data based on confidentiality.

show abstract

“…Bag-of-Words (BoW) is a classical mid-level feature representation framework which has demonstrated excellent performances in computer vision tasks such as image classification [13], [14] and action classification [15]- [17]. However, the traditional spatial pyramid widely used in BoW [13], [14] does not consider body structure information which is important prior information for person re-identification.…”

Section: Introductionmentioning

confidence: 99%

Orientation Driven Bag of Appearances for Person Re-identification

Ma,

Liu,

et al. 2016

Preprint

Self Cite

View full text Add to dashboard Cite

Person re-identification (re-id) consists of associating individual across camera network, which is valuable for intelligent video surveillance and has drawn wide attention. Although person re-identification research is making progress, it still faces some challenges such as varying poses, illumination and viewpoints. For feature representation in re-identification, existing works usually use low-level descriptors which do not take full advantage of body structure information, resulting in low representation ability. To solve this problem, this paper proposes the mid-level body-structure based feature representation (BSFR) which introduces body structure pyramid for codebook learning and feature pooling in the vertical direction of human body. Besides, varying viewpoints in the horizontal direction of human body usually causes the data missing problem, i.e., the appearances obtained in different orientations of the identical person could vary significantly. To address this problem, the orientation driven bag of appearances (ODBoA) is proposed to utilize person orientation information extracted by orientation estimation technic. To properly evaluate the proposed approach, we introduce a new re-identification dataset (Market-1203) based on the Market-1501 dataset and propose a new reidentification dataset (PKU-Reid). Both datasets contain multiple images captured in different body orientations for each person. Experimental results on three public datasets and two proposed datasets demonstrate the superiority of the proposed approach, indicating the effectiveness of body structure and orientation information for improving re-identification performance.

show abstract

A novel hierarchical Bag-of-Words model for compact action representation

Cited by 15 publications

References 46 publications

Review of the application of machine learning to the automatic semantic annotation of images

Review of the application of machine learning to the automatic semantic annotation of images

Classifying confidential data using SVM for efficient cloud query processing

Orientation Driven Bag of Appearances for Person Re-identification

Contact Info

Product

Resources

About