ggr2342

As a programmer where do I get started in computer vision and image processing?

Any kind of resource would be helpful.

There are important papers and then there are nicely written important papers. For the field of AI, ML and DL which papers are of the second kind?

Important works that are terribly written have to be filtered by a textbook author to take them to the masses. A nicely written paper doesn't require that.

Is this generally true or just my anecdotal experience. I have encountered physics, math, geology majors who learned to program because they needed it or they liked the idea of it. Later these were the people who turned into awesome programmers rather than someone who exclusively studied CS in undergrad.

Did anyone else have this experience?

Everyone has some pivotal moment in their career that propels their career in a positive direction. If you are a programmer or software engineer, what was that pivotal moment that made your skills increase in folds?

The aim is to understand books on proof based calculus, analysis, linear algebra, probability, etc.

Mathematical logic has confusing things like exchanging position of quantifiers changing meaning of entire statements, etc.

Where to learn the language of mathematics (suitable for self study) so that I can read other serious math books and topics?

I have dabbled with many resources before in hope of learning to program and learn basics of CS. I have looked at intro course sequences of schools like MIT, CMU, Stanford, etc. Most have partial resources locked down. I have dabbled with books and they felt like shit. Even the book Think Python by Allen Downey was all over the place.

The books and blogs at least taught me basic syntax constructs like variables, function definition, loops, etc. But I couldn't understand how to package them up (compose) to use them in solving problems.

Then I started studying through Berkeley's 3 course intro CS61A, CS61B, CS61C. They have all materials in the open and if you are following the most current iteration of the course, they even post solutions to the problems. It was godsend for me.

Now, I am confident to learn more CS topics using courses from CMU, MIT, Berkeley, Stanford, etc. The thing that was holding me back was the lack of confidence of programming and understanding really what a program was doing.

The field can by any field and by expert I don't mean the number 1 world expert. By expert I mean someone who can confidently do work (create things) in their domain for quite some time. Some fields are so vast there can't be a single top most expert.

If you consider yourself someone like that and assuming you read (past and present) books of that field, which books have you found the most influential for your knowledge and expertise?

Peter Norvig's pytudes is a great example of what I am talking about. Are there other such code repositories or similar things in any language that one can learn from?

Data wrangling is often considered not a beautiful task, to put it nicely. But the skillset is an immensely important one, according to me. It opens new doors in places.

Python is a good scripting language for that purpose. SQL is also helpful (AFAIK). Where to learn all about data wrangling? Are there books, courses, tutorials?

What other tools or languages can one learn for the purpose?

Which math book(s)/textbook(s) were the most valuable to you knowledge and skill development wise? In a sense, making even more math topics much more easier for you to understand.

Like, for me, Spivak's Calculus and Hubbard and Hubbard's Linear Algebra + Multivariable Calculus literally gave me wings. I began visualising problems in terms of matrices and transformations. Later I read Jaynes Probability Theory with much more ease. Because I got into thinking mathematically.