Activity recognition through multi-scale motion detail analysis

Du, Youtian; Chen, Feng; Wang, Xu; Zhang, Weidong

doi:10.1016/j.neucom.2007.09.012

Cited by 26 publications

(18 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…7 V is index for the angle rotation of the ellipse. 8 V is the index of horizontal position of the salient point of the convex.…”

Section: P V H P H V P V P V H P V H P V H P V H P V H P V H P H P H mentioning

confidence: 99%

“…At the lower level, they adopted HMM to recognize atomic actions and the output provided service for the higher level with context-free grammar mechanism. Du et al proposed HDS-DBN (Hierarchical Durational-State Dynamic Bayesian Network) to model interactions between two persons [8]. They decomposed an activity into multiple interactive stochastic processes and each corresponding to one scale of motion details.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A detection method based on Bayesian hierarchical network for abnormal interaction

Su¹,

Song²

2016

Proceedings of the 2016 4th International Conference on Advanced Materials and Information Technology Processing (AMITP 2016)

View full text Add to dashboard Cite

Abstract. Detecting the abnormal human interactions is vital in our daily life, especially when the society pay more attention to public security. But most researches didn't spare enough attention on abnormal interactions. In this paper, salient features are extracted for abnormal interactions, and the amounts of features are reduced to decrease the computation burden. Based on the extracted features, Bayesian hierarchical network is applied to estimating the pose of both persons. Then the corresponding rules for abnormal interaction detection are proposed. Finally, detection results are achieved based on the rules. UT-Interaction dataset is used for experiments. And the results show that the method outperforms with other methods in precision and sensitivity.

show abstract

“…7 V is index for the angle rotation of the ellipse. 8 V is the index of horizontal position of the salient point of the convex.…”

Section: P V H P H V P V P V H P V H P V H P V H P V H P V H P H P H mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A detection method based on Bayesian hierarchical network for abnormal interaction

Su¹,

Song²

2016

Proceedings of the 2016 4th International Conference on Advanced Materials and Information Technology Processing (AMITP 2016)

View full text Add to dashboard Cite

show abstract

“…Among most popular are hidden Markov models (HMMs) [10,11], support vector machines (SVM) [1,8,12], decision forests [8,13], Gaussian process dynamical models [14], k-means clustering [2] or dynamic Bayesian networks [15].…”

Section: State Of the Art In Gesture Recognitionmentioning

confidence: 99%

Full body movements recognition – unsupervised learning approach with heuristic R-GDL method

Hachaj

Ogiela

2015

Digital Signal Processing

View full text Add to dashboard Cite

“…Motion features have been used in different applications like video retrieval [32], action recognition [33], and human identification [34] to name a few. Periodicity of motion and its rate are the key property of obscene videos, which can be used as a cue for the feature extraction and classification.…”

Section: Features Based On Motion and Periodicity Characteristicsmentioning

confidence: 99%

Content-based obscene video recognition by combining 3D spatiotemporal and motion-based features

et al. 2012

View full text Add to dashboard Cite

In this article, a new method for the recognition of obscene video contents is presented. In the proposed algorithm, different episodes of a video file starting by key frames are classified independently by using the proposed features. We present three novel sets of features for the classification of video episodes, including (1) features based on the information of single video frames, (2) features based on 3D spatiotemporal volume (STV), and (3) features based on motion and periodicity characteristics. Furthermore, we propose the connected components' relation tree to find the spatiotemporal relationship between the connected components in consecutive frames for suitable features extraction. To divide an input video into video episodes, a new key frame extraction algorithm is utilized, which combines color histogram of the frames with the entropy of motion vectors. We compare the results of the proposed algorithm with those of other methods. The results reveal that the proposed algorithm increases the recognition rate by more than 9.34% in comparison with existing methods.

show abstract

Activity recognition through multi-scale motion detail analysis

Cited by 26 publications

References 20 publications

A detection method based on Bayesian hierarchical network for abnormal interaction

A detection method based on Bayesian hierarchical network for abnormal interaction

Full body movements recognition – unsupervised learning approach with heuristic R-GDL method

Content-based obscene video recognition by combining 3D spatiotemporal and motion-based features

Contact Info

Product

Resources

About