DEV Community

Hardik
Hardik

Posted on

Resources to get started with Computer Vision

Great resource for resources and pathway related to how to start with Machine Learning

Good Courses

Anndrew Ng Coursera ML basic ML course : if you are totally new to machine learning (else you can skip it)

Stanford CS229 : Haven’t myself done it yet, but looks very cool for those who are just starting with ML in general

3B1B’s series of neural nets : Very helpful for building a strong base and intuition

Stanford CS231n Youtube Course & Its website : Very good resource to start with, clears out the basics of ML in the first few videos along with building an intuition of CV

                                                                 OR
Enter fullscreen mode Exit fullscreen mode

Michigan University’s EECS 498 : Almost the similar content but explained in depth and in an easier fashion ( by the same instructor )

3B1B Convolutions Video : Before starting off the CNN in CS231n, I’d do this first, since it gives you a very good base and intuition of what Convolutions in actual math mean and then go back to 231n since I feel it starts on a bit of coarse note this way.

Topics which should be known if you wanna be called a good CV Research

  • Basics of ML
    • Bias Variance
    • Classification
    • Regression
  • Strong foundation of Neural Nets ( very clear intuition of how they work exactly )

  • Image Processing

    • Convolutions

      • Gaussian Blur, Mean Blur, Sobel Operator (for detecting edges)

        FOR DETECTING EDGES : Black and white filter ⇒ Gaussian Blur ⇒ Sobel (both in X and Y direction and then square rooting over the sum of their squares )

        since, it will remove the unnecessary edges from the images and just keep the main ones which are important for the image

Top comments (0)