great introductory stuff, great way to keep in touch with tensorflow's new tools, and the instructor is absolutely phenomenal. love the enthusiasm and the interactions with andrew are a joy to watch.
A really good course that builds up the knowledge over the concepts covered in Course 1. All the ideas are applicable in real world scenario and this is what makes the course that much more valuable!
автор: Prabesh G•
Okey.. So easy but okey
автор: Tanguy C•
Thanks. Enjoyed it.
that's was awesome
автор: Patrick L•
I like this course
автор: Vivek S•
Super cool stuff!
автор: Paulo A C•
автор: ashraf s t m•
Voice is low
автор: Venkatesh S•
автор: Bingcheng L•
автор: Hamzeh A•
автор: Omar M•
автор: S. M S H•
автор: Anastasia C•
I had a problem with the weekly assignments 1 and 4, which I think asked for things that were not presented at all in the videos. On the first week, without any preparation, we were asked to create the directories, supposedly without any python background, which was tough. But that was just a couple of commands. In the last weekly assignment, the file reading and loading was very problematic for those who hadn't previous python knowledge, and pretty advanced too. In the course of the lessons and the notebooks shown previously, never had we encountered .csv files, and the way to load the images that we were introduced to, with directories, was not at all present in the final project. Also, the methods that were needed afterwards (.flow(), evaluate) were not even hinted before, even in the comments of the assignment, if not before, during the videos/notebooks. All in all, the last exercise took me by surprise and was really tough to get working, because the course was almost irrelevant to it (no transfer learning or directories, the two main points in the course).
A bit too basic and shallow in terms of conducting the lecture. You are left doing most of the things on your own as the trainer assumes you know. Like using the jupyter notebook, configuring the tensorfow. Some of the google collab books do not work or took too long to load, the videos are too short no notes provided at all. After finishing the course there is nothing to refer to and its starting all over again. Given the level of machine learning course with Professor Adrew Ng, the standard is very high and you will expect that same level. Nevertheless, the concepts are very useful and the lecture explain very well. There level of material left for students to practice on their own,like assignments, notes. Not to be referred to existing material.
автор: Muthukumarasamy S•
Overall learning from this course is less compared to the expectations from a 4 week course. I was expecting to learn variety of TensorFlow implementations for CNN like Face recognition, Object detection. But this course only talks about Image classification. It would have been better if you could also discuss more about implementing various architectures in TensorFlow like ResNets, Inception. Also, You talked only about using sequential layers in Keras and concatenation of layers in Keras is not discussed here. I know all these concepts are discussed in Deep Learning specialization. I was only expecting to learn their implementation in TensorFlow from this course.
автор: Pablo A•
It's a nice next step after the first course in this series, however, I think a lot of this could be summarized in a shorter course or even added to course 1. I was particularly annoyed by some of the assignments as they required knowledge of other libraries that are not part of the course. Particularly Week 2 and 4, I spent a lot of time figuring out how different libraries worked just so I could preprocess my data before even gettin on to the course material. Week 4 in particular feels cramped up and the assignment uses a lot of tools not previously discussed, I don't think I learned much from it, I just wanted to be done.
автор: Dhruv D•
I wanted to rate this course 5 stars as it really is the best intro to CNN's ive seen but the last assignment was so egregious that I just have to dock 2 stars to bring this to instructor attention. Lots of people are having this problem. It was not the difficulty (in fact it was a nice change of pace from the usual flow_from_directory assignments) but instead the marking criteria and timeouts etc. My first submission took 2.5 hours to fail. I submitted again with next to no changes and it passed within 5 minutes - I lost the time it couldve taken to do .5-1 week of the next module
автор: Jesus E R•
I think it is too basic and not a lot of depth into what you are learning, specifically on the differences between binary and categorical classes, reading from disk vs manipulating data already in an array.
I was expecting more depth on tradeoff within Tensorflow's API choices. More direct comparisons between optimizers and the data generators. I feel like one straightforward exercise at the end of a really shallow week of videos is not enough to understand what students are really doing/learning.
The course lectures are solid, but the assignments are pretty dismal for beginners. There isn't much guidance built into the assignments, and sometimes they require the use of things that were absolutely not covered in the lectures(classic academic mistake). My suggestion for the course creators is to examine how Andrew Ng's assignments are in his Coursera course and model them after that. Or simply make sure that the assignments are clear(clear to someone beginning, not a TF expert).
автор: Artem D•
I liked the lectures (videos). And I did not like that the course has no mandatory programming assignments. I pay for the course to make myself study. And I believe that there is no study without practice. Hence, this course did not make me study, thus I don't understand why I need this course :-(. And I could find free lectures about TF/Keras (maybe not so good, but free) and/or read the documentation. BTW, I really like Andrew NG's courses, but this one really disappointed me.
автор: Shehryar M K K•
This course focuses on the teaching of TensorFlow modules related to CNNs and does a good job in introducing some modules of tf and keras for data loading and manipulation. However, it is very light on theory and is only helpful if Deep learning specialization is taken beforehand or in conjunction. Furthermore, this course will need some refresh soon for its modules as it is still using version v1.x of tf as well as some code re-organization.
автор: Zhuang L•
The videos were quite solid. The programming assignments were poorly designed to accept identical answers, but not other solutions that work. This did not evaluate students' creativity and depth of understanding. The Jupyter notebook environment was quite fragile. The resources allocated for each notebook was quite limited. I expect more computer or human resources allocated for each student paying the tuition.
автор: Thomas B•
This course teaches you how to apply CNN to image data, how to augment image data with ImageDataGenerator, and how to do transfer learning. It is very easy to follow, and quite possible to finish in half a days worth of effort. It would be nice to be more explicit with what is required by the grader, as assignment instructions not always are clear.
автор: Bakhtawar U R•
Good but too basic.
Specialization's first course already covered the basic of tensorlfow. This course is suppose to expose to sota topics in computer vision using cnns. The content in this course can be easily fetched from many online forums. Thus the curators need to put some advance topic like attention, spatial transformer etc etc