Great content, intuition, and examples. Some of the logistics and technical components were a bit more glitchy than the previous courses in the specialization - but that's a very high bar.

The course content is very good. I faced some problem in the assignment section. Even though in all the sections I was getting the expected answers I was not able to pass the assignment.

A very comprehensional lesson about convolutional netwoks but there are some little problems of the programming exercises that some of the contents may not be right and not fixed on time.

The course is great, it is a shame that the last assignment has a problem and I had to waste a lot of time trying to figure out what was happening until I found the problem in the forums.

Great course for general concept understanding. Assignments are well structured although could be slightly more deep with lesser hints. (currently it seems easy with all the hints given)

I think they have presented the most possible content that doesn't make course boring or useless. I learned new things which learning them takes a considerable amount of effort and time.

excellent course and very challenging!! The last programming assignment was a little frustrating having the correct answer and not getting credit for it until I changed the axis to None.

A great course with many simple code samples using tensorflow and Keras. The only issue was that instructors do not participate in the forums. Otherwise the content quality was excellent

Great course but in comparison with others of the specialization, the exercise did not leave that much as before.

Too focused on lines and not on the whole implementation of the ConvNet

The content is awesome, but using the old version of tf is not the best and obscures from the objective of either understanding the mechanics completely or creating CNNs with freedom.

Andrew Ng's simple explanation of complicated concepts lets you grasp concepts that require hour-long reading of research papers in a few minutes, Assignments could be more elaborate.

Der Kurs hat mir sehr geholfen und war Zeitlich gut zu schaffen.

Bei den Programmieraufgaben hätte ich mir ein paar beispiele gewünscht, um schneller in die Syntax einsteigen zu können.

It can be so much better if Andrew covers some more overviews. But it is still a great course and everyone should take it. Andrew gives the most understandable lessons about Yolo ones.

The first week was good and well presented. From the 2nd week onwards it as very cryptic. In the final week, on what are deep conv nets are learning, more details could have been given

I feel some of the theory was rushed here and high level functions used instead of explaining some of the more difficult concepts in the papers referenced, such as the Inception paper.

Good course, gives you a rough overview of popular CNN architectures and the basics of how they work. Assignments are easy but give you a chance to apply some of the stuff you learned.

The content was great, as usual, from Prof. Ng. But I wasted a lot of hours trying to make the programming assignment submissions, since auto save in IPython Notebooks kept on failing!

Minus one star for the assignment issue (triplet loss) for Week 4, this is a bit beyond acceptable considering how long the issue has evidently been there. Otherwise, stellar content.

Excellent content and explanations as usual. A small criticism for this course compared with the previous three is that the videos have clips of repeated content due to strange cuts.

Course content was good,but grader needs to be updated as in many cases I had to submit assignments manually by uploading it,as the grader wasn't working.Everything else was good. :)

Weird grader error on the week 4 assignment in triplet_loss function -- had to put axis=None instead of axis=-1, which is wrong (and contrary to the directions) -- needs to be fixed.

Great course. Very good and simple explanations of complex topics. The 4 is just for the sake I hoped for a deeper dive into Tensorflow and how everything changes with Tensorflow 2.

The problem with the triplet loss function and the grader was not solved until Dec 1st. This was misleading me, and, in order to grade fully I wrote a wrong version of the function.

A clear presentation of difficult problems and I love working with the notebooks. However meanwhile most of the tensorflow documentation hyperlinks within the notebooks are broken.