This week you'll learn about Enhanced Super-Resolution Generative Adversarial Networks (ESRGAN).

Image

The deep learning (DL) community has the appearance of a huge singular unit moving in only one direction: forward. Hence, it shouldn't come as a surprise when the already successful Super-Resolution Generative Adversarial Networks (SRGANs) were taken one step further with the introduction of Enhanced Super-Resolution Generative Adversarial Networks (ESRGANs). Nevertheless, the community strives to show that there is always more to be done. 

Super-resolution in itself is a domain that is slowly reaching its saturation point. However, with the introduction of ESRGANs, the DL community gets another reminder that there are always a million ways we can focus on improving the heart of our research and coming up with better results. 

The big picture: ESRGANs work as successors to SRGANs by keeping the core foundation but introducing just enough additional reinforcements and changes to make the whole concept better and more efficient. 

How it works: ESRGANs bring in updates like scraping the use of batch-normalization layers in the generator, the improved perception loss, pixel loss, etc. 

Our thoughts: ESRGANs are extremely intuitive, and the updates are easy to understand. It is a great topic that definitely shouldn't be overlooked. 

Yes, but: The issue of requiring extreme computing power and data still persists, disrupting the hopes of a normal ML practitioner (with standard machinery) recreating the paper's results. 

Stay smart: Try this out with your own dataset. Conversely, experimenting with the current dataset and the additional functions we have created can bring stark changes. There are a million possibilities, and it is up to you for the direction you would like to go! 

Click here to read the full tutorial

Solve Your CV/DL problem this week (or weekend) with our Working Code

You can instantly access all of the code for Enhanced Super-Resolution Generative Adversarial Networks (ESRGAN) by joining PyImageSearch University. Get working code to

  1. Finish your project this weekend with our code
  2. Solve your thorniest coding problems at work this week to show off your expertise
  3. Publish groundbreaking research without multiple tries at coding the hard parts

Guaranteed Results: If you haven't accomplished your CV/DL goals, let us know within 30 days and get a full refund.

I want the code


The PyImageSearch Team
Image


P.S. If you're interested in learning how to successfully apply deep learning to your own projects, I would recommend reading my book, Deep Learning for Computer Vision with Python.

Inside the book you'll find:
  • Super-practical walkthroughs that present solutions to actual real-world image classification (ResNet, VGG, etc.), object detection (Faster R-CNN, SSDs, RetinaNet, etc.), and segmentation (Mask R-CNN) problems
  • Hands on tutorials (with lots of code) that show you not only the algorithms behind deep learning for computer vision but their implementations as well.
  • A no-nonsense teaching style that is guaranteed to help you master deep learning for image understanding and visual recognition
If you're interested in learning more about the book, I'd be happy to send you a PDF containing the Table of Contents and a few sample chapters:

Click here to grab the PDF of sample chapters and Table of Contents

After clicking the above link, you'll receive a separate email with the PDF in a few short moments.