I really have to thank Ritesh-Megha and Maurya here. I had my MBA entrance exams to give and was occupied with the preparation for that most of the time. As far as reading up for the project, to understand the paper, the statistics, etc. I didn’t do any of it. Invariably it would be Maurya who would have read up the stuff on Google, and would explain it to me so that I wouldn’t be groping in the dark when the three of my team members had highly technical discussions about the project. I am sure they must have harbored plenty of anger for me as I shunned my responsibilities. They never communicated that anger, through any channel. I was already working under unhealthy levels of stress, and had maurya, ritesh and megha chosen to force me to do the work, my mind would have exploded to smithereens, scattered all over Parshavnath campus. In my defense I did take the ownership of the programming part and learnt MATLAB to make up for my poor participation in the first half of the project.

With that load off my chest, lets continue with the story. We were assigned our project guides. And there were mixed feelings regarding the guide we got. We were pretty aware of the fact that the word “guide” was a misnomer for any professor tagged to us. But let’s just go ahead with the term and avoid the complications of coining a new profile to describe the actual role played by the professor who was supposed to help us in the project. All we wanted from our guide was to accept the project as we gave it to him, not make a lot of fuss about the details and give us good marks. We were not sure if our guide would live up to the expectations.

The project was going along at a painfully slow pace. We were told that we would have to make a presentation at the end of the 7th semester. That got us moving. But understanding the project and the nitty-gritty of the entire face recognition was taking time.

It would be pertinent to explain the project in brief here. The face recognition technique we were using was quite easy to understand in principal. We represented every face as a combination of a fixed set of faces. For example, my face is 30% brad pitt’s face, 50% shahrukh khan’s face and 20% sachin tendulkar’s face. Now these percentages are stored in the computer. Also, the percentages would be unique to every face. So now, when a new face comes along, the percentages are calculated by the system, compared with the ones in the database, and if there is a match, the face is recognized, else the system will say it’s an unknown face. Now, it’s not nearly as simple as picking up sachin, shahrukh and brad pitt and finish the project. To make it sound complicated, and to try get through the point that we actually did work on our project, we had to make eigen faces from a set of standardized faces. After that we were supposed to represent every face to be stored in the database, or the ones on which recognition process was to be executed, in terms of percentage of these eigen faces. For all of that, we first had to extract the face from the picture a camera would capture. So we had to write routines for background subtraction, skin tone detection and optimize those routines to ensure that the face was accurately captured. Then we had to scale the face to a standard size to eliminate the differences in the size of the face in each picture. And to end it, we had to make a user friendly GUI (graphical user interface) so that a user could use the software. Whew!!! We did do a lot of stuff there. And I have not even explained what does an "Eigen" face mean!

For the seventh semester, we decided to restrict ourselves to creation of eigen faces. The code for that was pretty simple. Most of the effort went to the actual understanding of the process to carry out the face recognition. There was a hell lot of maths involved. After understanding that we had to implement that on Matlab. That was later. One week before the presentation was due, we sat down to prepare for it. We brain-stormed, finalized the scope, made the presentation slides, rehearsed the presentation, sitting for long hours alternately at my and Megha’s place. Maurya was quite aggravated by it all considering he had to reach all the way to thane after our technical picnics. Well, I am sure it must have been a good rehearsal for the job he got an year later which required the same late night journeys back home from Sukurwadi bus depot.

The day of the presentation dawned. Dressed in our formal attire, the class gathered in the auditorium room. After a series of presentations, our turn finally came. I was managing the slide transition as Maurya first stepped on the stage to deliver the opening and the introduction. An impeccable performance! After that, the plan was that Ritesh and I would explain the mathematics, and the implementation on MATLAB respectively, alternately. He would explain a little math, I would explain the corresponding implementation. That required us both to be on the stage simultaneously. We thought it would be smart, as well as different.

But then, plans have a way of going awry at PCT. The variables are just too many. And the slightest trigger can cause an avalanche that will leave your mind smothered and clouded. During our presentation, this trigger was provided by Ritesh. We were all too aware of his impulsive behaviour. But even we didn’t see the act of authority that Ritesh conjured up on the stage, presenting with me. As we were presenting, Don, the prof for whom the filmy line, roughly translated to English is “his friendhip ain’t good, neither is his enmity”, was casually chatting away with some other prof. Ritesh’s pride was terribly wounded, or so he gave the picture, as he thundered at Don to keep quiet. And the whole scene, repeated itself again, giving a sense of a nightmarish déjà vu. Once bitten, and bitten again, Don was not going to be shy to convey who was the boss. For all his intellectual short-comings he was still our prof. and by that virtue we were to respect him. Even if we didn’t respect him, publicly humiliating him was definitely not allowed. Ritesh had done just that. I was screaming in whispers, as loudly as I could without being overheard beyond the stage, imploring Ritesh to keep quiet. I glanced at Maurya. His expressions didn’t say much, but I am sure he was more frozen than liquid helium could have frozen him. By then, Don had obviously decided to screw us up bad. First thing he objected to was Ritesh and me presenting together. That was promptly resolved as I left the stage before Ritesh could raise any objections. By then Ritesh’s system had taken in what his emotions had got the group into. He tried to make amends by trying to pass a friendly smile in Don’s direction. But it was a lost battle. Ritesh finished his part and then I followed it up and then Megha, answering all the questions that were shot at us, which all seemed to try and exact a reason to flunk us. Thankfully, our understanding of the subject was sufficiently exhaustive to answer the queries of the professors. The questions started getting stupider by the minute and we kept answering. Everyone was convinced that we had worked on the project, and the implementation was on the way as well, and the project was good, but that didn’t mean that we would pass. Personal grudges could weigh out all the factors for which we should have got the highest marks. And there was more than sufficient reason for Don to bear a grudge. Like a company that goofs and than expends all its resources in a hectic and visible PR activity to be seen positively by its consumers, Ritesh too started his PR exercise. For whatever reasons, thankfully we passed. And thankfully Don never flunked Ritesh in any exams or vivas. Don certainly could have pulled some strings and given Ritesh, and us, KTs at will. I guess we ought to be thankful to Don for keeping our last year clean.