Temporal video segmentation by event detection: A novelty detection approach

Krishna, Mahesh Venkata; Bodesheim, Paul; Körner, Marco; Denzler, Joachim

doi:10.1134/s1054661814020114

Cited by 8 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, classification of such algorithms can be done in various ways. Anomaly detection can be achieved by segmenting a video in temporal domain, where each segment can be classified into a different category of interest [3,40]. In such algorithms, the overall change in scene dynamics within any chosen time interval is considered as an important feature towards identifying a temporal segment of interest.…”

Section: Related Workmentioning

confidence: 99%

Motion anomaly detection and trajectory analysis in visual surveillance

Chebiyyam

Reddy

Dogra

et al. 2017

Multimed Tools Appl

View full text Add to dashboard Cite

Motion anomaly detection through video analysis is important for delivering autonomous situation awareness in public places. Surveillance scene segmentation and representation is the preliminary step to implementation anomaly detection. Surveillance scene can be represented using Region Association Graph (RAG), where nodes represent regions and edges denote connectivity among the regions. Existing RAG-based analysis algorithms assume simple anomalies such as moving objects visit statistically unimportant or abandoned regions. However, complex anomalies such as an object encircles within a particular region (Type-I) or within a set of regions (Type-II). In this paper, we extract statistical features from a given set of object trajectories and train multi-class support vector machines (SVM) to deal with each type of anomaly. In the testing phase, a given test trajectory is categorized as normal or anomalous with respect to the trained models. Performance evaluation of the proposed algorithm has been carried out on public as well as our own datasets. We have recorded sensitivity as high as 86% and fallout rate as low as 9% in experimental evaluation of the proposed technique. We have carried out comparative analysis with state-of-the-art techniques to benchmark the method. It has been observed that the proposed model is consistent and highly accurate across challenging datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

Motion anomaly detection and trajectory analysis in visual surveillance

Chebiyyam

Reddy

Dogra

et al. 2017

Multimed Tools Appl

View full text Add to dashboard Cite

show abstract

“…The method proposed in [19] is unsupervised in nature, and a oneclass classification (OCC) technique is used based on Gaussian process regression (GPR) [19]. Table II shows that our method does not detect extra segments (γ = 0) for all videos except V1.…”

Section: Action Proposals Qualitymentioning

confidence: 99%

“…In Table II, we summarize the comparison results of our method with [12] and [19] based on detection rate η using (5) and over segmentation ratio γ using (6) on CCD dataset. The method proposed in [19] is unsupervised in nature, and a oneclass classification (OCC) technique is used based on Gaussian process regression (GPR) [19].…”

Section: Action Proposals Qualitymentioning

confidence: 99%

“…For evaluation, we also used detection rate η and over segmentation ratio γ [19] given by: number of True Positive Propsals number of segments in ground truth   (5) number of False Positive Proposals number of segments in ground truth…”

Section: A Datasets and Evaluation Measurementioning

confidence: 99%

“…Recent methods produce either spatiotemporal proposals using tube-let [14], the action-ness measurement [17], action tubes [15], segments based upon dense trajectories [12] or temporal proposals using e.g. fast activity proposals [18], the bag of fragments [13] and Gaussian process regression (GPR) [19] based methods. For proposal generation these methods either use hierarchical grouping methods or the dense trajectories of motion fields, which are computationally expensive for large-scale videos [18].…”

mentioning

confidence: 99%

See 2 more Smart Citations