Sorry, you need to enable JavaScript to visit this website.

NEAR-DUPLICATE VIDEO DETECTION EXPLOITING NOISE RESIDUAL TRACES

Citation Author(s):
Silvia Lameri, Luca Bondi, Paolo Bestagini, Stefano Tubaro
Submitted by:
Luca Bondi
Last updated:
19 September 2017 - 5:25am
Document Type:
Presentation Slides
Document Year:
2017
Event:
Presenters:
Luca Bondi
Paper Code:
WA.L4.1
Categories:
 

Video phylogeny research about joint analysis of correlated video sequences has shown the possibility of developing interesting forensic applications. As an example, it is possible to study the provenance of near-duplicate (ND) video sequences, i.e., videos generated from the same original one through content preserving transformations. To perform this kind of analysis, accurate detection of ND videos is paramount. In this paper, we propose an algorithm for ND video detection and clustering in a challenging setup. Specifically, we analyze a scenario in which many videos, depicting the same event, are recorded by different users. This situation is critical as non-ND videos acquired from very close viewpoints run the risk of being incorrectly detected as ND. The proposed approach leverages on robust hashing properties and the concept of sensor noise traces.

up
0 users have voted: