So, first of all, how did I come up with this idea for my Final Year Project?
Actually I didn’t. At the end of October I received an email from Prof. Alan Smeaton about FYP and decided to take one of his prescribed projects.
I had to do loads of research as I had no clue about images, their features and how other image annotation techniques work. All I knew was that this is an interesting project and I want to do it no matter how much I need to learn to complete it. The idea of meeting people from the INSIGHT Research Center, and learning from professionals also had an impact on this decision.
I met with my supervisor few times before preparing first delivery component which was Idea Proposal Documentation. This document included topics such as Project Concept, what I expect to deliver, some market analysis and proposed timeline. This involved some research and helped me to get a clearer picture what exactly will need to be done. Few weeks later I have delivered a Pre Proposal Presentation during which the idea was approved. Very important lesson I have learnt from this ‘Don’t assume HDMI cable is something that’s used everywhere and it will be available’. So, to not get surprised again, I’ve got all possible adapters and cables with HDMI which I will bring everywhere with me from now for connecting my laptop.
After official approval of the idea I could start to prepare my Initial Delivery Documentation which covers areas such us Project Summary (Concept, Users, Market Size), Expected Outcome, Project Timeline, Prototype Functionality, Technologies Used, Summary Functional Specification (End User Requirements, Architecture Diagram) and Technical Challenges. I found some of the sections easier to write about than others, and some still need more details and some are unknown because of the research nature of this project.
In the meantime Prof. Alan Smeaton has introduced me to the team of researchers from the Insight Research Center, who will work on the research project involving the same dataset that I will be using. I will attend some of their meetings as it may help me with development of my prototype. These meetings are to be held in January and I hope I will find answers to some of my questions that will help me to finalize my Initial Delivery Documentation.
The following are two links that describe the dataset and show its potential:
The article that can be accessed from this link proves that a need for good image annotation technique is still not fulfilled and that it is a very relevant area of research: http://www.nytimes.com/2014/11/18/science/researchers-announce-breakthrough-in-content-recognition-software.html?_r=0