 Hello everyone. My name is Atharva Morone. I am from VNIT, Nagpur. The basic objective of my project is to create an assessment tool set for painting. So this is the overview. My software basically compares two images as everyone else's. The teacher's painting and the student's painting. And grades the image according to its close nets with the teacher's image. While doing this, basically I faced many challenges. So I'll come to challenges. So I'll first come to challenges. So there are two types, non-technical and technical. The first is the non-technical in which I faced the challenge, evaluation of free form of painting. See there are two types of paintings. First is the free form where there are no well-defined boundaries and the colors interact with each other in complex ways. So use of technology to appreciate their aesthetic beauty is one of the major challenges that I faced. And hence it was decided that we will stick with only a sketchbook type of thing where the student is already given the outline. He just has to paint in it and then upload the image. The second challenge is finding painting course material online. Now this is a challenge because most of the painting courses that I searched online were paid. So their material could not be accessed directly. Moreover, they were not in the form which were required by us to make such kind of an application. So that was again a challenge. Then third is deciding the parameters for grading. Now this was a challenge because first thing is we don't have much domain knowledge of painting. The second thing is that due to the lack of domain knowledge and use of technology to assess that painting. So both this club together became a challenge. So that is why this is a challenge. Then technical challenges are the following. First is compensation for different image capturing environments. Now this is a major challenge. See when a student uploads an image he can do it in various ways. He can do it in white light. He can do it in sunlight. He can do it in even a shadow. I mean under a tree or something like that. He can do anything. He has the freedom. He is a user. This will make the changes in color itself, the image captured. So the colors that are captured will be a little different from the original drawing. So first that has to be compensated otherwise it will give fault results. The second thing is rotating and scaling the input image so that it matches best with the standard image. Again as Saurav had also said, this was again a challenge because there are many standard algorithms in the open civil library but none of them worked as expected and we even tried writing our own software but none of them worked for all the test cases. So again that remained as a challenge. So due to all these challenges some of them are still unresolved and due to these the application is not yet completed but I'm sure if more research is given and more time is given it will be completed in the future. Thank you.