Space-Time Zernike Moments and Pyramid Kernel Descriptors for Action Classification

Costantini, Luca; Seidenari, Lorenzo; Serra, Giuseppe; Capodiferro, Licia; Bimbo, Alberto Del

doi:10.1007/978-3-642-24088-1_21

Cited by 10 publications

(10 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The presented set is able to eliminate the transformation factors, no matter in translation, scaling, reflection, skew and rotation. Since then, the research topic of moments function has been extensively explored for the past few decades (Almoosa et al, 2008;Chen et al, 2013;Costantini et al, 2011;Li et al, 2012). Every publication has reported its improved version of moments.…”

Section: Feature Descriptor Formulationmentioning

confidence: 99%

Visual Tracking using Invariant Feature Descriptor

Ong¹,

Lau²,

Koo³

2017

American Journal of Applied Sciences

View full text Add to dashboard Cite

The process of identifying the state of an object in a video sequence is referred as visual tracking. It is mainly achieved by using the appearance information from a reference image to recognize the similar characteristics from the other images. Since a digital image is built-up with rows and columns of pixels that are represented with finite set of digital values, the appearance information is measured with a mathematical formulation that is known as image intensity. The problem of distinguishing the intensity of the object of interest from the other objects and the surrounding background is always the main challenge in visual tracking. In this study, a novel invariant feature descriptor model is introduced to address the aforesaid problem. The proposed framework is inspired by the theoretical model of local features that has been widelyused for image recognition. From the large number of diversified scenarios in the surveillance applications, the performance of the proposed model is demonstrated with the benchmarked dataset of single-target tracking. The experiment results shown the advantage of our proposed model for tracking non-rigid object in the changing background as compared to other state-ofthe-art visual trackers. In addition, the important aspects of the proposed model are analyzed and highlighted as well in the experimental discussions.

show abstract

Section: Feature Descriptor Formulationmentioning

confidence: 99%

Visual Tracking using Invariant Feature Descriptor

Ong¹,

Lau²,

Koo³

2017

American Journal of Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Gao et al propose spatial-temporal volumes. 19 Costantini et al 20 adopt multiscale interest point cluster. Inspired by these approaches, we propose a new feature representation structure: super-interest point.…”

Section: Related Workmentioning

confidence: 99%

Human Action Recognition Based on Normalized Interest Points and Super-Interest Points

Wang

2014

Int. J. Human. Robot.

View full text Add to dashboard Cite

Visual-based human action recognition is currently one of the most active research topics in computer vision. The feature representation directly has a crucial impact on the performance of the recognition. Feature representation based on bag-of-words is popular in current research, but the spatial and temporal relationship among these features is usually discarded. In order to solve this issue, a novel feature representation based on normalized interest points is proposed and utilized to recognize the human actions. The novel representation is called super-interest point. The novelty of the proposed feature is that the spatial-temporal correlation between the interest points and human body can be directly added to the representation without considering scale and location variance of the points by introducing normalized points clustering. The novelty concerns three tasks. First, to solve the diversity of human location and scale, interest points are normalized based on the normalization of the human region. Second, to obtain the spatial-temporal correlation among the interest points, the normalized points with similar spatial and temporal distance are constructed to a superinterest point by using three-dimensional clustering algorithm. Finally, by describing the appearance characteristic of the super-interest points and location relationship among the super-interest points, a new feature representation is gained. The proposed representation formation sets up the relationship among local features and human¯gure. Experiments on Weizmann, KTH, and UCF sports dataset demonstrate that the proposed feature is e®ective for human action recognition.

show abstract

“…3 This second approach is especially employed in the human actions recognition algorithms. [4][5][6][7][8] These algorithms are very important in many computer vision applications such as video surveillance or human computer interfaces.…”

Section: Introductionmentioning

confidence: 99%

“…These functions were used in the context of the shape retrieval 13 and, recently, in the context of human actions recognition. 8 In this work we make use of two descriptors to summarize the space-time patches. Both of the descriptors are based on the Zernike polynomials but the main differences between them is their symmetry, indeed the first one is based on a spherical symmetry, and the second one is based on a cylindrical symmetry.…”

Section: Introductionmentioning

confidence: 99%

A comparison between space-time video descriptors

2013

Self Cite

View full text Add to dashboard Cite

The description of space-time patches is a fundamental task in many applications such as video retrieval or classification. Each space-time patch can be described by using a set of orthogonal functions that represent a subspace, for example a sphere or a cylinder, within the patch. In this work, our aim is to investigate the differences between the spherical descriptors and the cylindrical descriptors. In order to compute the descriptors, the 3D spherical and cylindrical Zernike polynomials are employed. This is important because both the functions are based on the same family of polynomials, and only the symmetry is different. Our experimental results show that the cylindrical descriptor outperforms the spherical descriptor. However, the performances of the two descriptors are similar

show abstract

Space-Time Zernike Moments and Pyramid Kernel Descriptors for Action Classification

Cited by 10 publications

References 20 publications

Visual Tracking using Invariant Feature Descriptor

Visual Tracking using Invariant Feature Descriptor

Human Action Recognition Based on Normalized Interest Points and Super-Interest Points

A comparison between space-time video descriptors

Contact Info

Product

Resources

About