Preparing for the autumn trick (vision algorithms)

    Autumn trick, I'm going to prepare the basis for projects and papers in three aspects. Provisional basis "Introduction to Algorithms" + LeetCode, internships and projects currently I am the only complete set of project, and finally intensive thirty-four papers.

basis:

        First, the establishment algorithm, the classic basic algorithm do not know it seems a bit unreasonable. I dream in their own way no matter how much is not clear, these things always will, and I do not want to sleepwalk do not understand anything of it had finished in this life. The basis of it, even less than the future, at least to training thinking, broaden the ideological boundaries to help solve problems when it opens up new ideas.

        Secondly, if the above can get these done (fear is not possible) within one month, you learn some basic aspects of the image depth of learning, such as listening to a class at Stanford and the like.

        However, to my understanding of their own, the above content ranging from completion, there will be new ideas emerge, want to learn something new. . . East West, a hammer and a stick of the cycle until the time is wasted. So here we want to do something serious plan.

        1. look at the directory and then reading or lectures. Page after page one a bow looking back, I do not know what the ultimate goal, get lost on the way, finally gave in, "this book with nothing too hard do not want to see" in the sense that the process is very boring boring.

        2. Basic has a goal plan. The goal is you want to learn these things for its own purposes, which part is more difficult to plan, how long it takes to read the title will do, which is part of the previously studied but weak, weak points where, how to know by heart.

project:

        The project will certainly be the focus of the interview.

        Three questions learn from others: it is a kind of project? What technology is used (very detailed clearly and logically that)? Encountered any problems, how to solve?

        First out technical part:

        1. fold clothes generation:

  • Generating a random walk + Bezier curve
  • Mobile is the Least Squares
  • Adjust the brightness

        2. Magnets spots generated:

  • Supra, the transition region color simulation balabala
  • Rust generation (randomly generated single communication region)

intensive reading

        Since it is intensive, you do not have to do not so new, classic and career planning related read by heart.

  • BOTH
  • Resen
  • Wgan
  • FasterCNN

Guess you like

Origin blog.csdn.net/qq_26271435/article/details/93319722