Deep learning for computer vision

This is one of the few reviews I ever wrote, so please excuse my style, not really my turf :)

So why then write it ? Because this book is that good that I felt it really deserves one.

I got in touch with Adrian Rosebrock’s blog quite a while ago, when I was tinkering with OpenCV, RPI and computer vision. Since the first time I landed on his blog ( www.pyimagesearch.com ), it has immediately become my go-to source whenever I got stuck in issues with Python and/or OpenCV.

Then I took his course in OpenCV. And then he published this wonderful book.

I have a penchant on taking courses on several sites ( coursera, udemy, skillwise and a few others ), and all I can say is that Adrian made it from start in my top 1% instructors. The combination of clarity that he reaches in explaining complex topics with the attention to detail he has in implementing anything is fantastic and hardly matched.

Now, a few words about the book. Due to the vast horizon that he covers, it’s split in three parts, starting from the basics of how neural networks work, going through very detailed explanations on CNNs, and then of specialized architectures like FasterCNN, Yolo, and a few other, and completing with hardcore topics like distributed training. In my opinion, the bar was already high based on his previous works, but he surpassed it with brio. This book is now my textbook for computer vision.  

So, if you’re in search on how to start in this fascinating domain, or you just want to improve your skills, head over to DL4CV and start your journey :)

Thanks, Adrian!