A survey on using domain and contextual knowledge for human activity recognition in video streams

Onofri, Leonardo; Soda, Paolo; Pechenizkiy, Mykola; Iannello, Giulio

doi:10.1016/j.eswa.2016.06.011

Cited by 79 publications

(32 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can effectively soften its hard constraints and deal with uncertainties while compactly representing many relationships. First-order logic is a knowledge base consisting of a series of sentences or rules [Onofri, Soda, Pechenizkiy et al (2016)], and the Markov logical network gives weights to each rule, softening the hard rules. From the perspective of probability, Markov logic network can be flexibly and modularly combined with a large amount of knowledge.…”

Section: Event Recognition Based On Logical Reasoningmentioning

confidence: 99%

Research on Action Recognition and Content Analysis in videos based on DNN and MLN

Song

Zhao

et al. 2019

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

In the current era of multimedia information, it is increasingly urgent to realize intelligent video action recognition and content analysis. In the past few years, video action recognition, as an important direction in computer vision, has attracted many researchers and made much progress. First, this paper reviews the latest video action recognition methods based on Deep Neural Network and Markov Logic Network. Second, we analyze the characteristics of each method and the performance from the experiment results. Then compare the emphases of these methods and discuss the application scenarios. Finally, we consider and prospect the development trend and direction of this field.

show abstract

Section: Event Recognition Based On Logical Reasoningmentioning

confidence: 99%

Research on Action Recognition and Content Analysis in videos based on DNN and MLN

Song

Zhao

et al. 2019

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

“…Evidence garnered from cognitive science research shows that humans require semantic information, such as the context, scene, or interacting objects, to recognize actions (Biederman et al 1982). In many studies in the computer vision field, attempts have been made to include semantic information in human activity recognition (Ziaeefard, Bergevin 2015;Onofri et al 2016;Herath et al 2017). Marszalek et al (2009) exploited the context of natural dynamic scenes for human action recognition in video clips.…”

Section: Related Workmentioning

confidence: 99%

Enhancing action recognition of construction workers using data-driven scene parsing

Yang

2018

Journal of Civil Engineering and Management

View full text Add to dashboard Cite

Vision-based action recognition of construction workers has attracted increasing attention for its diverse applications. Though state-of-the-art performances have been achieved using spatial-temporal features in previous studies, considerable challenges remain in the context of cluttered and dynamic construction sites. Considering that workers actions are closely related to various construction entities, this paper proposes a novel system on enhancing action recognition using semantic information. A data-driven scene parsing method, named label transfer, is adopted to recognize construction entities in the entire scene. A probabilistic model of actions with context is established. Worker actions are first classified using dense trajectories, and then improved by construction object recognition. The experimental results on a comprehensive dataset show that the proposed system outperforms the baseline algorithm by 10.5%. The paper provides a new solution to integrate semantic information globally, other than conventional object detection, which can only depict local context. The proposed system is especially suitable for construction sites, where semantic information is rich from local objects to global surroundings. As compared to other methods using object detection to integrate context information, it is easy to implement, requiring no tedious training or parameter tuning, and is scalable to the number of recognizable objects.

show abstract

“…They can be grouped as "statistic, syntactic, and descript-based methods". The public and private sets of data used in this frame video can be helpful for readers in weightlifting training (5).…”

Section: Previous Weightlifting Workmentioning

confidence: 99%

“…Human action in video frames is a strong research field that offers some applications in the computer visions, such as netting-based control, contents-based video analyses, etc., which are used with the increasing number of cameras (3), (5).…”

Section: Introductionmentioning

confidence: 99%

Proposing an Analysis System to Monitoring Weightlifting Based on Training (Snatch and Clean and Jerk)

Journal¹

2018

Baghdad Sci.J

View full text Add to dashboard Cite

Analysis system of sports players is very important for individuals in weightlifting. Assessment of player and strength is important for the performance of weightlifting. This paper proposes an analytical method for weightlifters with check-by-frame video. This analysis system can compute the major steps of seven positions in both snatch and clean and jerk methods in frame-video weightlifting monitoring of movements. Each user can compute the major steps of the seven positions of Hu moments among two frames in the video during training, and the Euclidian distance can be computed for the Hu moment values and lifting moment values in the snatch and clean and jerk methods during training. The outcome of the proposed system shows on efficient, accurate results in monitoring movement analysis in weightlifting for playersduring training in this area.

show abstract

A survey on using domain and contextual knowledge for human activity recognition in video streams

Cited by 79 publications

References 46 publications

Research on Action Recognition and Content Analysis in videos based on DNN and MLN

Research on Action Recognition and Content Analysis in videos based on DNN and MLN

Enhancing action recognition of construction workers using data-driven scene parsing

Proposing an Analysis System to Monitoring Weightlifting Based on Training (Snatch and Clean and Jerk)

Contact Info

Product

Resources

About