Firstly, the Coursera submission system for the final project is a disgrace, and that's why I removed one star.

Other than that, the course is amazing. By far the best in the series, and I don't understand the complaining. At this level, the instructor does what he needs to. Go through the needed functions and explain their use and inputs.

The final assignment is also great and gives very good knowledge, and more or less all that is needed is already in the lectures. You just need to copy/paste, out them in order and go through the documentation a bit to correct mistakes and understand what's missing.

I think the issue is the first 2.5 courses that are ridiculously easy and offered more or less everything already solved, without challenging students to start learning and thinking properly. You cannot expect people to learn by answering all practice questions in the previous video and the graded assignments can be passed by c/p the same code and changing a single variable...

Great course, at the correct level. Really glad the specialisation ended with it.

The gist of the course is excellent, but challenging - the instructor doesn't hold your hand through every step this time. You're pointed at some relevant Python library modules and given an overview of the basics for them with examples, but you'll have to spend a lot of time reading documentation and figuring out what to do with them. That's okay, because as Professor Brooks points out throughout the course, the idea is for you to get a taste of how you'd need to work going forward.

This is a new course, so there are still a few rough spots, but it's being actively corrected as problems surface, and I am confident in saying that if you follow the course and do all the work, you'll have developed some hands-on experience!

To be honest the test of week1 was really frustrating. This course is totally different then the 4 previous courses. Course 1-4 with Steve and Paul was more like riding a bike with training wheels.

Then on course 5 Christopher came along and kicked them away :-) Felt quite frustrating. But at the end this is more like in real live, where you have to find the solutions on your own.

The lectures are ok. The video when you just watch someone reading the text didn't help much.

Even though i had some difficulties getting through the assignments, as they were not very clear with the objectives, i still learned a great deal and i can say i became better at python. Learning new libraries was a great experience and getting to know the extent of python was an experience in itself.

i would have rated it 5 star if not for very poor submission system.

this course is very challenging and if you are not up for challenges think twice before enrolling.

The pier grading system is quite slow and working on the notebook with opencv was a bit hard because the kernel kept crashing.

I think the course projects are really good. The projects require you to learn new modules with minimal help, which is excellent for those who want to become a real programmer. The ability to learn individually is invaluable to any programmer, and this course did provide good training on self-learning.

The problems with this course are also obvious. For example, there are quite a few errors in the jupyter notebook setup video for MacOS. Those errors might not be a big problem, but for people who did not have previous experience in setting up a virtual environment, the typos in captions could cause problems. That particular video did tell us what we need to do, but it just missed some details, and that made things a bit confusing.

The next problem is with Coursera's peer review system. In the final project, we are instructed to paste our code to the Coursera textbox. However, the textbox is not code-friendly at all (it messes up all the indentation and spaces), and Coursera won't let us copy the codes other people submitted. These problems made it very hard to review other people's work. I am disappointed about it because I think the peer review session is, in fact, a good chance for us to learn from other students. The Coursera system is not making it easy for us to read and copy codes from others, and the instructor did not give us a proper workaround instruction in the submitting page. All these problems could actually be avoided if the instructor suggested a better submit option.

Last but not least, I hope the instructor could make the videos more engaging than it is right now. Honestly, he did not add much value with those videos. He was reading the text on the jupyter notebook most of the time...I don't need someone to read the text out loud, I can read the notebook all by myself.

All in all, I believe the people who designed this course need to put more focus on the details. It is a shame that a capstone course is unnecessarily made more difficult than it should be.

I really liked the content and its presentation. The knowledge conveyed in this course is useful, the teaching staff does a good job, and the final project is legitimately interesting. If you aim only for knowledge and don't care about the certification, go aheas, you will not regret it.

The reason this is not 5 stars is the very, very end - the submission procedure for the final project. It involves pasting Python source code into a Coursera text field. This kills the Python formatting, and thus the code becomes unrunnable. Workarounds have been suggested, but did not work for me (and at least a few others).

It is completely beyond me why this method of submission was chosen. The course creators themselves use a way better method earlier in the course (html export of Jupyter notebook).

Why is this worth a 2 star deduction? Because it means that people may pay for this and aim for a certification, and could fail at the very end for reasons they have zero control over.

There is a big gap within previous courses from this specialization and this one, even though the videos are clear, the pace is really fast. Practically they just show some libraries with a few examples and then they just leave you alone with the final project. Thanks to Google I was able to finish the project, I learn a lot doing it. One thing that is horrible is the Jupyter notebook on Coursera, very s....l....o....w, and they don't facilitate to download the project on a easy way, again using google you can get to something like this: https://www.reddit.com/r/learnmachinelearning/comments/7er5ps/coursera_downloading_all_the_assignments_jupyter/ on my case I had to split the file on 90 MB and then download all the parts On my machine was really fast to complete the project, I spent more time train to complete the project directly on Coursera and then searching solution to the speed issue than actually writing the final project

The lectures and labs in this course are informative, but the format could and should have been much better. Some examples for potential improvements include shorter and better structured lectures, quizzes and a few additional short assignments (similarly to other courses in this specialization). The final project was very time-consuming, in such a way that did not make much sense. The task was to guess some parameter settings that lead to a specific pre-defined, but barely acceptable result for face detection. In this case, an improvement could have been to specify a minimum number of faces to be detected in each image for a passing grade (while keeping the number of false positives at 0 or below a certain threshold), and give extra points for further correctly detected faces. Overall, I have learned a lot in this course, but the gain for the time investment could have been larger with a better prepared course.

This course might be good for some who are interested to learn PIL and OCV but for python language specialization this course is a bit off topic. They should have taught advanced programming concepts after classes and inheritance e.g. idiomatic programming. This course is taught at a high speed too. Sometimes it becomes hard to follow. May be in future they add more relevant content related to language instead of this course, in the specialization. The focus of this course is to teach how to get help, use and read documentation from different python libraries, which do not required a separate course to my point of view. Neither this course covers pillow, tesseract, and opencv concepts in detail.

Michigan please review the content in this course and make it more relevant to python programming, I know you can do better. See you in another specialization.

The Peer-review project is not well designed. The instruction about what to submit is misleading. Students are confused about what to submit, uploading a PDF like before, or pasting the whole HTML file?

This becomes clear only after you submit and start to review peer's submission. Students are supposed to copy and paste their code into the submission window. But there is a BIG problem of this approach. The pasted code lose all the formatting, and python relies on formatting. Even if I want to copy a peer's codes and try to run them in Jupyter, the codes are all messed up. There is no way to figure out what my classmate wants to do.

The labs provide all the script of the lecture, so the lecture videos are pretty much useless. The projects in this course are unexpectedly more difficult than previous courses, and the teaching assistant seems to be complaining about students' posts all the time, making me hesitate to ask questions there. I give 3 stars only because I spent a lot of time and efforts and I felt that I learned a lot by the end of the course. But still, this course could be designed a lot better than this. I will not be surprised that many students just give up in the middle of the learning.

There are significant problems with processing and submitting the capstone project; the Coursera Jupyter notebook runs out of memory when working on a particular file (U Mich, you might want to simply reduce some of the image sizes). In addition, the submission process doesn't work without linking outside to a google drive or something of the kind.

The information, though, was excellent, and I'm happy I've completed the whole specialization. Course 3/5 for the reasons listed above (they're quite frustrating) specialization 5/5.

I feel the pace of this course is a bit too fast, and it is hard to master all materials learned about the three major APIs (tesseract, openCV, and pillow) given the lecture. A lot of self-experimenting and trial-and-error is required to do the final project. A more through introduction to each of these three excellent API would be helpful and reduce the burden of the final project. Nevertheless, I sincerely appreciate the efforts the instructors devoted to this specializations. I learned a ton from each of you. Cheers!

This course was a real disappointment when compared with the other courses in the Python 3 Programming specialisation. Steve and Paul were EXCELLENT in the first 4 courses, building concepts up slowly and assigning challenging but manageable final assignments that integrated everything learnt prior. This capstone on the other hand had a complete 180 in terms of pedagogy. If the aim was to teach students to learn how to dig through messy documentation, fiddle with libraries and dependencies, and read page after page of stackoverflow forums, then a simple project requiring us to find a library and explain several features would’ve sufficed. Instead, we got three libraries thrown at us with nary an explanation of how to use them. I’m not talking about simply reusing code that has already been constructed by Christopher, but rather, going through analytical skills of how to deconstruct any library and work out the functions. I’ll still recommend the course, but the final capstone does leave a bad taste. I felt like I was tracking to hack together some abomination, instead of purposefully constructing a tool step by step. A shame given that the incremental style of development was emphasised by Steve and Paul a lot. Here, I was adding lines of code without care as to what they did, just in the hopes that it would give me the outputs needed to pass.

"Capstone Projects are hands-on projects that let you apply what you've learned in a Specialization to a practical question or problem related to the Specialization topic."

Capstone Projects – Coursera Help Center

In this capstone course I expected to apply what I had learned in the previous four courses. I had already taken Python for Everyone, when this course utilized Python 2. I wanted a course about Python 3 which included classes, inheritance, etc. The use of the runestone textbook is great!, but I wish that it had included more about classes, inheritance, etc. I expected that the capstone would be an application in which I was required to apply a class structure with inheritance, etc to a problem.

This capstone was about opencv, pytesseract, and pillow, which are worthwhile topics, but these topics are complex and should be covered in a separate specification. To learn Juypter, opencv, pytesseract and pillow within three weeks is asking too much for students who are MOOC learners (parttime).

I rated this capstone two stars, because the capstone missed the stated purpose of a capstone according to Coursera Help Center.

While the topic of this course and the libraries that it covers are extremely interesting and worth studying. I believe that they were not discussed on a deep enough level in order to come to the total realization of what's going on inside of them.

Generally being advertised as a part of python specialization for beginners as well as people without technical beckground, this particular course reaches far outside a scope of knowledge, which could have been achvired though the previous courses of the specialization.

Withouth providing enough materials to practice, it leaves you with a bit of theory how certain libraries work and then throws at you grands projects, which additionally require extensive research from your side in order to come to means with all the tools you are to use in the process.

My opinion is that this particular course is not suitable for beginners. Even after completing all the previous courses from this specialization.

Even though the pillow, tesseract and opencv libraries are very cool and interesting to use - the project does not present a useful challenge. The instructor felt like he was just reading code off a script and there was no interactive problem-solving or debugging.

The project's scope was too specific in its usability and the explanation of the libraries was also lacking. It leaves too much for you to figure out on your own - which would be great if it was challenges related to writing good code, but the challenges were related to figuring out how to use these specific libraries and spend time finding methods which let you do what you wanted it to. I think the project can be improved by giving more clearer instructions on the libraries, but more challenge on the programming logic.

Given the nature of image processing, a lot of time was also spend waiting for the code to finish running before you can debug (and each run took minutes)

It's very disjointed from the previous 4 courses, as it's taught by a different instructor with a very different style.

I get it that the goal is to get learners out of the simple & clean bubble of the learning environment and deal with messy real-world one. But from the lecture videos, it's clear that the instructor makes assumptions of students' Python skills and just skim through many concepts and codes expecting students to understand them. This doesn't make sense. If these concepts are utilized building towards the capstone, why not have them elaborated in an extra course before the capstone? The learning experience is terrible as I feel I've missed out many things left unexplained (despite Googling).

I'll suggest having instructors from the previous courses in this specialization to do the capstone course for consistency purposes.

For those who take this course with the completion of the specialization in mind, take note that the difficulty increase quite drastically for this course as compared to the other courses in this specialization. Also beware that this course has a more figure it out yourself rather than a guide approach to teaching and the lecture has a more read the script approach to it, at least we get to read the script for revision and play with the lecture code. Some may be out off by it, but i would say that this is better than other capstones or courses in Coursera, at least the instructions for coursework are clear and the coding environment for it work.... as compared to other more buggy approach in other courses... but beware the Jupyter thingy takes a long time to run your code...

Awful, especially when compared to the other 4 courses in the Python 3 Programming specialization. Minutes on end of the teacher showing how to hit the help function and then partly reading through it is definitely not what any student should get upon paying for a course.

Other reviews of this course have gone into enough detail on how unintuitive it is to submit the projects, be sure to read them before enrolling in this course.

A final point: the vibe in the discussion forum is a huge downer, and it seems to be because of the hard rules enforced with very little empathy towards students.

As a beginner I found it quite difficult as compared to other courses in this series. I was expecting a lot more handholding and detailed explanation of the subject as it was done in the other courses in this otherwise great series. However I really enjoyed learning.

A dismal course, which is a disgraceful "effort" by:

"Christopher Brooks is an applied Computer Scientist who builds and studies the effects of educational technologies in higher education ..."

Esoteric and unintersting information that doesn't tie back to anything done in the previous courses for this specialization.