Hello trend,
Today, we are sharing a post on Super Resolution using OpenCV.
People are not impressed when you show them the state of the art in computer vision. The reason is that computer vision competes with science fiction, and we are always behind. In the movies, they simply have to "enhance, enhance, enhance" to zoom infinitely into eyes of the murder victim to see the murderer clearly. Mystery solved.
So when computer vision researchers came up with super resolution, no one is impressed. Sigh!
Anyway, today we have post on super resolution where you take a picture and ENHANCE (ahem) its resolution. Now, you may be thinking how are we able to get the details that are not present in the image to begin with? The short answer is that we train a neural network that learns how to upsample an image based on millions of images.
In this post, we go over four different models EDSR, ESPCN, FSRCNN, and LapSRN. They have their own pros and cons but the good news is that you do not need to know the theoretical details to use the models. It is less than 10 lines of code (and a pre-trained model) right in the post. Follow the link below to learn more.
https://learnopencv.com/super-resolution-in-opencv/
OPENCV WEEKLY WEBINAR
I am thrilled to announce that we are starting a weekly webinar series at OpenCV. It premiers in 3 hours - 9 AM Pacific Time!
In this series, we will cover a variety of topics of interest to the OpenCV community. At the same time, we will get an opportunity to interact with our community members live. Sometimes we will cover a live tutorial, and at other times we will bring in guests to discuss a topic of interest to our community.
Here are the details of the first webinar, which premieres this Thursday:
Topic : OpenCV AI Competition 2021
Description : We will discuss the selection process, and go over a handful of projects that blew us away. We will also discuss general trends in spatial AI.
Time : 9 AM Pacific Time, March 25, 2021
Hosts : Satya Mallick (CEO, OpenCV.org), Anna Petrovicheva (CTO, OpenCV.ai), and Brandon Gilles (Chief Architect of OAK).
Link : https://us02web.zoom.us/j/83231402214
Questions : After the first 40 mins, we will open up the webinar for Q&A.
Satya
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.