Information-based adaptive fast-forward for visual surveillance

Höferlin, Benjamin; Höferlin, Markus; Weiskopf, Daniel; Heidemann, Gunther

doi:10.1007/s11042-010-0606-z

Cited by 43 publications

(45 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Summarization can also be interpreted as a query process, where the playback speed is adjusted according to the similarity between the frame and the target content [29]. Adaptive fast-forwarding can be considered from the perspective of information theory, with the goal of equalizing the scene complexity, represented by the statistical distance (alpha-divergence) between the frame difference and the learnt noise model [10]. Various visualization techniques for fastforwarded summaries can be used [38].…”

Section: Related Workmentioning

confidence: 99%

“…Unlike previous methods that considered low-level features only, e.g., motion vectors [9] or frame differences [10], we consider video tracking and hotspot detection on surveillance videos, and the combination of player tracking with detection of camera motions and various production actions for processing broadcasted soccer videos.…”

Section: Use Casesmentioning

confidence: 99%

“…Especially, we compared the behaviour of our proposed method to three methods, i.e. Peker et al [9], Höferlin et al [10] and Naive fast-forwarding.…”

Section: A Experimental Setupmentioning

confidence: 99%

“…We estimated the motion vector by the Horn-Schunck method as originally applied by Peker et al, and used the implementation in OpenCV. Höferlin et al [10] determine the activity level by computing the alpha-divergence between the luminance difference of two consecutive frames and the estimated noise model. A bigger divergence value stands for a larger distance between the current frame difference and those caused by the background noise.…”

Section: A Experimental Setupmentioning

confidence: 99%

“…Conventional fast-forwarding-based methods mainly subsample frames based on information associated to scene activity, defined via optical flow [9] or the histogram of pixel differences [10]. By only evaluating changes in the scene, it is difficult to assure the semantic relevance of the summary.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Resource Allocation for Personalized Video Summarization

Chen

Vleeschouwer

Cavallaro

2014

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Abstract-We propose a hybrid personalized summarization framework that combines adaptive fast-forwarding and content truncation to generate comfortable and compact video summaries. We formulate video summarization as a discrete optimization problem, where the optimal summary is determined by adopting Lagrangian relaxation and convex-hull approximation to solve a resource allocation problem. To trade-off playback speed and perceptual comfort we consider information associated to the still content of the scene, which is essential to evaluate the relevance of a video, and information associated to the scene activity, which is more relevant for visual comfort. We perform clip-level fast-forwarding by selecting the playback speeds from discrete options, which naturally include content truncation as special case with infinite playback speed. We demonstrate the proposed summarization framework in two use cases, namely summarization of broadcasted soccer videos and surveillance videos. Objective and subjective experiments are performed to demonstrate the relevance and efficiency of the proposed method.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Use Casesmentioning

confidence: 99%

“…Especially, we compared the behaviour of our proposed method to three methods, i.e. Peker et al [9], Höferlin et al [10] and Naive fast-forwarding.…”

Section: A Experimental Setupmentioning

confidence: 99%

Section: A Experimental Setupmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations