YOLOvME: Solving the Parking Problem

Vladimir Panteleev believes that Machine Learning is a hidden passion of any developer. He runs the R&D department at Paralect. First, the team used ML approaches for internal tools in Paralect, then began working on commercial products using ML solutions.

When did you start using YOLOv5?

Vladimir and his team first tried YOLOv5 only a couple of months ago. Since then, they have been amazed by how simple Computer Vision development can be!

How did you decide to use vision AI to solve the parking problem?

The problem is a piece of cake! Vladimir found that he was struggling to find parking in the lot near his office. So, he decided to use computers to help solve the parking spot problem! A webcam is simply placed in the window of a building making it possible to track free parking spots 24/7.

What made you specifically choose YOLOv5?

Vladimir and his team have gone through dozens of solutions:

• Self-hosted
• Cloud-based
• Open source
• Proprietary

All of them have different "nuances,” such as poorly supervised learning, over-consumption of resources, API/UI designed for the aliens, not for humans, and astronomically high pricing. So, YOLOv5 was our favorite from the very beginning as it ideally fit our needs, as well as our partners'.

What aspects of YOLOv5 made it easy for you to work with?

No doubt, the ability to choose models! It allows you to use YOLOv5n/s/m/l/x for different problems and it's amazing. The friendly interface, documentation, and fair resource consumption make YOLOv5 a leader in the object detection domain today.

What are other challenges you are expecting to solve with YOLOv5 in the future?

We are ready for any challenge, so if you read this article and are interested in building a "cutting-edge" solution for open source or any other purposes, please get in touch!

And lastly, what would you like to recommend to someone who is new to AI?

Don't be afraid and try! Ultralytics has made Computer Vision very accessible and affordable with YOLOv5. 10 years ago you literally had to be a Ph.D., now such solutions can be implemented by any developer.

Vladimir’s goal at Paralect is to provide affordable cutting-edge solutions for all founders! But, that atmosphere makes all of us think that it's something unsolvable. They advocate the idea of building in public to show the rest of the world that everything is much easier than it seems! Vladimir’s team is writing a guide to help other builders to get started with computer vision. Check out Vladimir’s other projects on Twitter!

If you think your use case should be shared with the world and you want to tell what you are doing with YOLOv5, tag us on social media @Ultralytics with #YOLOvME for a chance to be featured.