Theme 3 - Research and theory (1/2)

The research journal I have chosen is called Multimedia and is a part of IEEE. The journal has an impact factor of 1.694 and I believe that it is relevant for media technology research as it publishes research articles within subjects such as multimedia user interfaces, creation and modification of multimedia applications and media creation integration editing and management, among other topics. The journals goal is to keep readers informed about technical trends and research within the subject of multimedia.

The article I have chosen from Multimedia, IEEE is “Face Matching and Retrieval in Forensics Applications” written by Anil k. Jain, Brendan Klare and Unsang Park in 2012. The article studies facial recognition approaches and looks at difficulties that exist within this field that affect the accuracy of the results. The article identifies four problems within facial recognition and try to solve how the accuracy of facial recognition within these areas can be improved. The problems are sketch recognition, aging/facial scars/marks, pose-/illumination correction and low resolution images. The research method is not explained in detail but the article states that the method chosen is a survey. A weakness within the article is the lack of explanation in regards to the method, e.g. in how they collected and analyzed the data.

The article suggests three actions for improving the accuracy of the results. First, entering images into a 3D simulator with a specific algorithm for facial aging to get an improved image of what a person might look like at a specific age. Second, the article suggests methods for image filtering on low resolution images in order to improve and enhance the quality. Third, the article claims that the accuracy of matching sketches to photographs is very low. In order to solve this problem the authors designed a facial recognition system that actually uses both the sketch and photographs together in the same program in order to get more results and hopefully a better accuracy. The problems and solutions are described in detail but the article does not explain how they collected the data and arrived at these conclusions which makes it difficult to see if the conclusions are a logical extension of the data.

In connection with the proposals, the authors of the article also warns that technology alone will not lead to more accurate results. The article argues that technology will still not always be successful when trying to find a match, which means that human interaction is still needed to interpret the result.

Briefly explain to a first year university student what theory is, and what theory is not.

The term theory can be described as explaining something and testing it through scientific methods. A theory tries to provide an understanding for how and why something is done in that particular way. Theory is not a simple list of data, variables, diagrams, methods or hypothesis.


Describe the major theory or theories that are used in your selected paper. Which theory type (see Table 2 in Gregor) can the theory or theories be characterized as?

There are five types of reasons for why theories are made; theory for analyzing, theory for explaining, theory for predicting, theory for explaining and predicting and theory for design and action. The major theory in the article I chose is design and action. When problems and weaknesses were identified the artic

Which are the benefits and limitations of using the selected theory or theories?

A benefit of the theory for design and action is that it gives instructions on how to do something. In the example of “Face Matching and Retrieval in Forensics Applications” the article offers states how to solve specific problems. However, limitations of this type of theory is that is does not necessarily provide detailed explanations about the problems or even the proposed solutions.le gave detailed suggestions of how these problems could be solved.