Home | Students | Course Content | Homework | Projects |Announcements
To get off the ground on your project, you need to do some readings in a particular area of interest. To this end, you have to write a review of a recent paper.
Your review should focus on how the paper solves an important problem. What are the other solutions and how does it improve on it, and on the specific results they obtain. These need to be expressed as a single webpage which should be in the area yourarea/reviewPaper/index.html if you use extracts or images from the original paper or from any other source, you must cite them.
Language MUST be original, and we will check your work for plagiarism with the Turnitin software. However, you are free to borrow reasonable amounts of text so long as you cite it as a direct quotation.
This constitutes the first step of your project, and will be worth about 8-10% of the project grade or 5% of the course grade.
The papers you selected are listed below. The pdfs are also available in the directory papers, and are also linked from the titles below. A bibTeX entry for each paper is in the file projectPapers.bib.
SUBMISSION: Webpage: Sept 4 midnight Presentation: Sept 6-8 (schedule TBA)Never read a paper from start to end at one shot. What you need to do is to read the abstract, a bit of the opening lines, and then flip through the paper, reading the figures and captions etc. and some paragraphs here and there. Dabble here, dabble there. Look up some of the main citations. Perhaps you may want to download some of these papers. Do this for some time, before starting to actually read.
You may get as much as 50-80% of the gist of the paper by just skimming. It follows Zipf's law - 70% of the effort goes in understanding the last 30% of the problem. And it can save a lot of time if you can quickly determine that this is a paper that you DONT want to read in detail!!
Then, sit with a pencil and start reading in earnest, marking items and their meanings as you go. To understand the ideas, you may need to look up some of the background papers also. If so, just skim them so that you get some basic idea; don't attempt to read them.
But DO LOOK UP some of the references; it always helps understand the paper better than just reading the one paper.
This applies to a lot of documents, not just papers.
Y8167 | EE | BHUWAN DHINGRA | Estimating scene typicality from human ratings and image features | Review | Turnitin Report |
Jayant Sharma | Turnitin Report | 42% |
Sneha Agarwal | Turnitin Report | 31% |
Karthik M | Turnitin Report | 24% |
Hemangini Parmar | Turnitin Report | 23% |
Agrim Gupta | Turnitin Report | 21% |
Bhuwan Dhingra | Turnitin Report | 17% |
Anuj Gupta | Turnitin Report | 13% |
Ruhi Dang | Turnitin Report | 9% |
Shubham Tulsani | Turnitin Report | 9% |
Apurva Gupta | Turnitin Report | 8% |
Vidur Kumar | Turnitin Report | 7% |
Asmita Bhattacharya | Turnitin Report | 6% |
Debosmita Chaudhuri | Turnitin Report | 5% |
Sonal Kumari | Turnitin Report | 5% |
Shorya S Roy | Turnitin Report | 4% |
Abhishek Kar | Turnitin Report | 2% |
Sumit Verma | Turnitin Report | 2% |
Pranjal Saxena | Turnitin Report | 0% |
Shantanu Saraswat | Turnitin Report | 0% |