This is very intensive and wonderful course on CNN. No other course in the MOOC world can be compared to this course's capability of simplifying complex concepts and visualizing them to get intuition.
Great course for kickoff into the world of CNN's. Gives a nice overview of existing architectures and certain applications of CNN's as well as giving some solid background in how they work internally.
by Akash M•
When i started out with deep learning, i found Course 4 to be the most intriguing part of this specialization. And i was not disappointed. I already knew the scope of CNNs, but to see them in work from up close was a treat. This course teaches you the fine intricacies of Convolutional Neural Network. It also showcases the working of some really famous models that were built in the last few years. I hope this course can be extended to include the applications of CNN in NLP as well. This course is a must for budding Deep Learning Researchers. I cannot wait to apply the learnings in real life.
by Ayush T•
Like the other courses of this series, this course is really good. In this tutorial I have not only understood how to implement things but I have also learnt what's the math behind those things. It is important at-least for me because it allows me to do more experiments with CNN's or in general Neural Networks. The thing which I like most about this course is its programming exercises.
I recommend this whole series to those people who want to learn some advance machine learning stuff like GAN, variational autoencoders and Reinforcement learning. This series will help as a strong foundation.
by Yilun Y•
Overall an awesome course, however, it somewhat lacks some important topics and models such as SSD, Faster RCNN, mask RCNN, etc which are even more frequently mentioned in literature and applied in real world projects. This course really sparked my curiosity and passion in deep learning, I actually learned the models mentioned before by reading the original paper and many useful blogs. This is a long but rewarding journey, I would also like to see more topics be covered in this course and let more people know how these state-of-art models work and how they really change the world.
by Xiang J•
I really like this course, because it not only taught me the exciting new topics that I always want to learn, such as object detection algorithm and neural style transfer, but also it gave a solid introduction to the concepts of convolution. The assignments are great, it is fun to do and it also helped me more concretely understand the materials of main course. As to further improve the course, may be it would be nice to build a whole end-to-end pipeline including training the main convolution model in car detection as I know in Google colab even public users have access to GPUs.
by Mukund C•
Loved it!! Loved it!! Loved it!! I wish there was a little bit more engagement from mentor side as well as updates on the coursework with the latest developments in the object detection field. I also wish that there were a little bit more involved programming exercises, maybe one in "training" where one has to label objects and "train" a neural net. One of the things that I missed in the course is an explanation of the Neural Network architectures and why they work - e.g. the VCCG-16 or Inception Network - for example. Maybe one has to read the papers to understand them?
by Shankar G•
This part of the CNNs course in DL was awesome and long enough. It started with foundations of CNNs, where the concepts of CNNs layers was made very clear. Programming assignments helped understanding the layering activation properly. The good part was DeepCNNs case studies explanation with its pros and cons, plus the practical advice for using ConvNets. Also this course provided few papers applications like object detection, face recognition and neural style transfer which was amazing. All the quizzes and programming assignments refreshed the concepts in a good manner.
by Mahmoud s m•
i hope we could implement every code from scratch , i mean that you don't do the heavy lifting for us and we start the code from the zero point no matter how much time or effort it would take us , implementing codes in the existing manner is great , but creating it and passing through all phases of the code like arranging the code , efficiency in programming , the steps of writing a certain function also the arrangement of all functions like(which before which) .All of this will help us gain better hands on programming ourselves . thx for the great course :D :D
This course covers the basics of convolutional neural networks , resnets, inception nets, yolo, style transfer, face recognition.The programming assignments mostly for yolo and face recognition is done with transfer learning , i think its only fair as they are computationally expensive to train.I am confident about all the materials covered in this course Andrew Ng as always breaks down the problem to the basics so you can understand them.Its a great course if you want to know and implement the well known computer vision problems with the well known algorithms.
by Alouini M Y•
This course helped me consolidate my computer vision knowledge. In fact, I had some prior experience but felt left behind given the current rapid advancements in the field of computer vision (thanks to deep learning mostly). The material is up-to-date and the assignments (especially the notebooks) are very pleasant. I have learned a lot of modern CV techniques: YOLO for image detection and localisation, style transfer, face verificiation with DeepFace, and many more. I recommend to anyone that is serious (or at least curious) about modern CV techniques.
by Jeffrey S•
I had a tough time on the programming exercises - mostly due to poor Python/Numpy/Tensorflow experience. I did find the material really interesting. The teaching style is great - much better than other courses on AI I've started. Andrew is terrific and pleasant to learn from. While totally different from the megastar CS50 (EdX) approach, he manages to make a complicated subject understandable. I have my list of subjects I need to go back and review, but I really feel like I've gotten a good perspective on the Deep Learning field from these courses.
by Luiz E d F M•
Excellent course in all aspects, both in terms of difficulty and depth of learning. Thank you to everyone involved in this project for providing us with learning and obtaining such rich and essential knowledge for the present and the future. Many thanks to tutor @paulinpaloalto for always being such a helpful, considerate person, with a high level of knowledge and charisma. Thanks also to Andrew NG for being such an excellent teacher and master of the subject, and for teaching us so sublimely and dedicatedly in every detail of the specialization.
by Jairo J P H•
El curso es muy bueno, particularmente estoy muy agradecido con COURSERA, por darme la oportunidad de hacer los cinco cursos de la Especialización en Deep Learning con ayuda economica y permitirme tener acceso a este tipo de capacitacion y certificacion. Muchas Gracias…!
The course is very good, particularly I am very grateful to COURSERA, for giving me the opportunity to do the five courses of the Deep Learning Specialization with financial aid and allowing me to have access to this type of training and certification. Thank you very much!
by Jennifer J•
Fascinating course with brilliant insights into how deep convolutional nets work, however it would of been far better had the instructor used coded examples of math like those from the papers with code website which makes it easier to understand and translate the math into code. However, the exercises are fascinating, fun and outright brilliant nonetheless! It's worth completing this to gain an insightful and eventually coded math understanding of concepts such as neural style transfer and facial recognition. This can never get boring!
by Martin K•
Andrews unique way of presenting complex theoretical concepts in a compelling and easy to understand manner was essential for my learning success. Attending this course was fun. Even though the programming assignments were pretty tough in this course (for me the toughest of all the courses in the deep learning specialization), I managed to complete this course in (my) record time. This might be mostly due to the understanding of the underlying mathematical concepts which were outstandingly well presented.
Totally recommend this course!
by David A G•
The course was excellent. I really enjoy Andrew Ng's courses: complex stuff made easy and lots of practical applications.
The only thing that I would try to improve is the time the staff dedicates to check the forum to solve student's questions. I personally got stuck at one of the quizzes and it was hard to find any clue that might help to understand the right answer. Also, some really interesting general questions on the forum were not replied by anyone. I'm sure some expert help on the forum would bring great value to the course.
by Marcel M•
For an engineering discipline, there is nothing better than employing the latest state-of-the-art techniques in solving real-life problems. That's the inherent value of this course the fact that you learn how Deep Learning is having an impact on so, so.. many, diverse areas such as Self-Driving Cars, Object Detection, Localization, Classification, Verification, Recognition and much, more. I highly recommend this course to anyone who wants to be an adept Deep Learning Practitioner. Kudos! Team DeepLearning AI. Keep up the good work!
by J K•
The best course (yet). A good balance between theory and practice, although the complete lack of TensorFlow and Keras fundamentals can be a bit frustrating. Additionally, the use of numpy operations (add, multiply and such) gave the impression that you'd correctly done a function assignment (the check values were OK), however, the grader failed to accept it as being correct (which was justified), however, an indication that it was incorrect (or some comments in the accompanying text) would've saved me 30 minutes of searching.
by Ahmed E S A H•
This course is very good. But i hope, after the course's weeks end, to add one more section to explain the recent publications and the most important challenges in the course field. In my opinion, this section will help the researcher to find a path to start research in course topic and try to find a new contributions that can help them specially if there are new master's or PhD students, they can figure out quickly where to start there research topics.
Thank you for your great effort and i hope i can learn more via Coursera.
by Asif M•
Its a very complicated topic and Andrew Ng and his team have made it very easy for us to learn the core concepts and easily do the programming exercises. Needless to say, we need to spend some additional time outside the course if you want to get a deeper understanding of the topic as well as learn more about the nuances of pre-processing and loading data/models abstracted away by the utilities as well as the detailed instructions in the exercises.
PS: The discussion forum is super helpful, especially when you need some help.
The course is a perfect balance between theoretical explanations, application in programming and tips that can be helpful if you intend to work with CNN. I had not seen CNN before, and I didn't feel lost at any moment. Every doubt I had was perfectly answered in the forum. You don't need much of an experience with TensorFlow or Keras to do the labs, which are accompanied by thorough explanations of what is required; on the other hand, there are "extra" tasks for people who want to go more into depth in each lab.
by Vincent F•
Overall a very good course for the instruction. Found only two omissions with the programming assignment notebooks. One was where a function expected a tensor but the parameter we were encouraged to provide was an array. Had to use a convert to tensor call. The other was a mismatch between the expected output block and the grader. This has been noted already but has not yet been fixed. But quite minor all in all.
Really liked the links to the academic papers that are the source of the models used. Thanks again.
by Maximiliano B•
In this module of the specialization, you will be familiar with several types of convolutional neural networks and how do they work in details. Compared to the previous modules, this one requires more time due to the complexity of the subject as well as the programming assignments that are more difficult. After this course you feel comfortable to read all the papers covered as references throughout the course . Moreover, Professor Andrew NG explains the content clearly and it is a pleasure to watch his videos.
by José L•
Needs a few corrections on the last week's assignment. Other than that great course. I recommend people to go deeper (no pun intended) in learning Tensorflow and Keras by self studying via other resources (books, videos, tutorials) since the programming material is too extensive to teach in a course like this which seems intended to master the basic concepts and the most important results in convnets. Thank to Andrew and the TAs for an excellent course. See you all in the Sequence Models and last course!
by Kai-Peter M•
Great course!!! The best online course I have ever taken! I enjoyed almost every day I participated in that course, really an educational treasure! It is so comprehensive and detailed at the same time. Due to the good presentation of the topics it was really understandable. The only thing I would wish for future participants: please make it easier to get the complete Jupyter notebook environments from the Coursera platform once completed. I spent a lot of time here - even after consuming the related blogs.
by Matthew B•
Great course. Brilliant overview of CNN with recent implementations. I understand the limitations of covering only so much material in 4 weeks. Wish the course could have gone deeper on training YOLO. I had to do this myself from the darknet website with some other tutorials. Something to consider, implementations of Unet and Mask RCNN may be even more useful for precise object masking/detection rather than bounding box in the future. May be worth mentioning these techniques as they develop further.