Self-Evaluation

Evaluating my own work is very hard for me to do in this case. On one hand, I am very frustrated that I was not only unable to produce the desired results that I set out to achieve, but I was also given an extension past the original deadline that still didn't culminate into a more successful final project. On the other hand, I feel very good about the progress I've made in understanding why I was unable to succeed and I feel confident that if I had the answers to my questions about the specific issues I raised in the earlier parts of this final write-up, then I could indeed finish this project successfully. Moreover, I feel that trying to implement a second image cutout system greatly increased my understanding about the image segmentation problem and how better to solve it. On the progression of this project, last week Friday (March 10) was a very troubling day for me in terms of how I felt about my lack of progress and remaining lack of understanding for why things weren't working. Part of this though was a general feeling of fatigue and exhaustion over my work on this project. Debugging code is the absolute hardest thing for me to work on because quite often I will spend several hours pursuing a particular direction that in the end will get me nowhere. This was the case of the last week of this project. It was both frustrating and emotionally exhausting to spend so much time with little to show for it. However, I blame this stress on myself and my failure to be further along with my progress earlier in the timeline of this project. I feel it took me too long to get to the debugging phase and this in turn led to the very undesirable situation of trying to fix something without giving myself enough time to step back and really try to understand what I was doing. On Friday I informed you that this project summary would be late, which once again is something I didn't want to do, but I now feel was immensely necessary. Stepping away from this project for a day gave me some breathing room to come back to it now with a fresh perspective. As a result, I now feel I understand my problem and what it would take to solve it. That alone gives me a much greater sense of satisfaction than I had last week. I would recommend to anyone that they build in specific days just for walking away from a project entirely and coming back to it with a clear head periodically to help re-stimulate the development process. The other recommendation I would make for myself and for others doing a similar project is to periodically search for other work that might be going on that relates directly to what you're doing. In my case, having the "Implementing GrabCut" paper early on probably could have saved me a lot of time in the beginning when I was just trying to interpret the original paper.