1

Paper Review

An Effective Approach for Video Copy Detection and Identification of

Misbehaving Users

By: S.Sujath, and G.Kalaiaras

LEIZA LINDA L. PELAYO

MIT Student

Professor:

Sept. 25, 2018

Version: 1.0

Journal Paper Review

 2

Journal Paper Review

An Effective Approach for Video Copy Detection and Identification of Misbehaving Users
The authors, S.Sujath, and G.Kalaiaras, discussed the result of their study as to detect whether the
query video frames are a copy of a video from the train video database or not the Auto dual Threshold is
used to eliminate the redundant frame, the SVD-SIFT features are used to compare the two frames features
sets points, the Graph-based video sequence matching is used to match the query video and train video,
and If the query video is matched then server identifies the unauthorized user and block that identified
user at particular time. I.e. based on timestamp.
An experimental method was use by the authors in order to detect video copy detection. They also
apply Ordinal Pattern Distribution (OPD) histogram proposed in as the ordinal feature. For the color
feature, they characterize the color information of a GoF by using the cumulative color information of all
the sub-sampled I frames in it.
The authors used the auto dual treshhold method to eliminate redundant video frames, select
keyframes. C1f denotes the first frame of Segment 1, C1l the last frame of the Segment 1; C2f denotes the
first frame of the Segment 2.
In using the SVD – SIFT Matching, the authors found out that a comparative study of the
performance of various feature descriptors showed that the SIFT descriptor is more robust than others
with respect to rotation, scale changes, view-point change, and local affine transformations.
Another method used for video copy detection is graph-based video sequence matching in a 6-steps
procedures. Time direction consistency and jump degree was also used in this method.
The authors also made use of the timestamp which is the time at which an event is recorded by a
computer, not the time of the event itself. In many cases, the difference may be inconsequential: the time
at which an event is recorded by a timestamp (e.g., entered into a log file) should be close to the time of the
event. This data is usually presented in a consistent format, allowing for easy comparison of two different
records and tracking progress over time; the practice of recording timestamps in a consistent manner
along with the actual data is called time stamping.
In discussing the timestamp, the authors identified some advantages such as: (1) Server will identify
misbehaving user in the proposed system that is based on timestamp and block that identified user at
particular time, and (2) Additionally give opportunity to all user’s i.e. user’s tries to login 3 attempts and
block that identified user at particular time.
It is good that the authors used particular formula in the experiment such as the OPD histograms
and cumulative histograms of each channels which led to the result that the total dimension of the color
feature is also 3×24=72, representing three color channels.
The authors observed that seven of the ten theories that we identified are native to the SE discipline.
At first glance, they said that this observation may seem surprising, however they think it is traceable to a
2004 finding of Glass et al.(2004) who describe how the fields of Computer Science (CS) and SE are inward
looking and do not borrow theories and concepts from other reference disciplines. First observation of the
authors was that such theories are good for informing the design of new empirical studies, which might
Journal Paper Review
 3

have the long term benefit of helping the research literature converge on a particular topic (or research
question), as Stol and Fitzgerald (2013) indicated in their studies. Second, borrowed theories can
contribute to the development of SE-specific frameworks potentially useful to empirical SE researchers for
organizing existing concepts from the literature, or in assisting in the development and evaluation of
knowledge claims, e.g. hypotheses. And third, borrowing theories from other disciplines can increase the
explanatory power of the theory – explaining the constructs and propositions of a theory.
The authors also presented a list of transformations used in the content-based copy detection task
which was shown in table 1. The explanation was very thorough that it gave complete details of the
different tasks performed during the study.
In showing the framework for video copy detection system, the authors used a flowchart indicating
the flow of the processes used in the content-based copy detection.
At the end of the article, the authors concluded that in detecting the copy video becomes finding the
longest path in the matching result graph are obtained. Suppose if the result of the frame is matched, i.e the
matching frame result is been verified from the database which is already been trained. It identifies the
unauthorized user and block that identified user at particular time. I.e. based on timestamp, again the same
user queries the same video then that particular user timing will increases.

Critique:
The article is quite informative though lacking some evidence showing the actual application of the
result of the study. The article is impressing that it was published in the International Journal of Innovative
Research in Computer and Communication Engineering.
The authors have credible qualifications since one of them is an Assistant Professor in
Dhanalakshmi Srinivasan College of Engineering and Technology. Another point is that publishing a
research in an international fora is quite difficult, as such these authors are very commendable for having
this research published.

Conclusion:
I found the article very useful as such it provided relevant information in content-based copy
detections for videos. It was comprehensive enough to show that certain process and computations can be
made use to show and provide a good result for this study.

Journal Paper Review