Action movies segmentation and summarization based on tempo analysis

Chen, Hsuan-Wei; Kuo, J.B.; Chu, Wei-Ta; Wu, Ja-Ling

doi:10.1145/1026711.1026752

Cited by 49 publications

(33 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An LSU contains a series of shots that convey a solid semantic meaning. Moreover, various video abstraction techniques [4]- [6] have been proposed to represent movie content in a compact manner, such as automatic summarization for action movies [7], [8].…”

Section: Related Workmentioning

confidence: 99%

“…For movie videos, the so-called logical story units [3] are determined through checking shots' visual similarity and the pattern of shot changes. Based on tempo analysis [2] and computational media aesthetics [19], we developed a system to automatically perform story segmentation for action movies [8].…”

Section: Story Segmentationmentioning

confidence: 99%

“…We compare the performance of story segmentation based on four approaches-The tempo-based approach [8], the one with global thresholds, the storyshed algorithm, and the storyshed algorithm with thresholds. As described in [8], movie tempo is calculated based on motion activity, audio energy dynamics, and shot change frequency.…”

Section: B Story Segmentationmentioning

confidence: 99%

“…As described in [8], movie tempo is calculated based on motion activity, audio energy dynamics, and shot change frequency. According to the tempo curve corresponding to a movie, our previous work [8] can be applied to determine the boundaries of stories in movies. We compare the tempo-based approach with the newly proposed social-based approach to demonstrate the superiority of the latter.…”

Section: B Story Segmentationmentioning

confidence: 99%

See 3 more Smart Citations

RoleNet: Movie Analysis from the Perspective of Social Networks

Weng

Chu

2009

IEEE Trans. Multimedia

Self Cite

149

129

View full text Add to dashboard Cite

Abstract-With the idea of social network analysis, we propose a novel way to analyze movie videos from the perspective of social relationships rather than audiovisual features. To appropriately describe role's relationships in movies, we devise a method to quantify relations and construct role's social networks, called RoleNet. Based on RoleNet, we are able to perform semantic analysis that goes beyond conventional feature-based approaches. In this work, social relations between roles are used to be the context information of video scenes, and leading roles and the corresponding communities can be automatically determined. The results of community identification provide new alternatives in media management and browsing. Moreover, by describing video scenes with role's context, social-relation-based story segmentation method is developed to pave a new way for this widely-studied topic. Experimental results show the effectiveness of leading role determination and community identification. We also demonstrate that the social-based story segmentation approach works much better than the conventional tempo-based method. Finally, we give extensive discussions and state that the proposed ideas provide insights into context-based video analysis.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Story Segmentationmentioning

confidence: 99%

Section: B Story Segmentationmentioning

confidence: 99%

Section: B Story Segmentationmentioning

confidence: 99%

See 2 more Smart Citations

RoleNet: Movie Analysis from the Perspective of Social Networks

Weng

Chu

2009

IEEE Trans. Multimedia

Self Cite

149

129

View full text Add to dashboard Cite

show abstract

“…Directors may control and present the rhythm by the techniques of montage (shot length), motion, and audio effects. We then adopt the formulation proposed in [14] to compute the rhythm elements.…”

Section: Aesthetic Field Extractionmentioning

confidence: 99%

Film Narrative Exploration Through the Analysis of Aesthetic Elements

Wang¹,

Cheng

Chen³

et al. 2006

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. In this paper, we propose a novel method for performing high-level narrative structure extraction of films. Our objective is to utilize the knowledge of film production for analyzing and extracting the structure of films. This is achieved by combining visual and aural cues on the basis of cinematic principles. An aesthetic model is developed to integrate visual and aural cues (aesthetic fields) to evaluate the aesthetic intensity curve which is associated with the film's narrative structure. Finally, we conduct experiments on different genres of films. Experimental results demonstrate the effectiveness of our approach.

show abstract