Day 85 – Real-Time Multi-Object Tracker Using YOLO v5 and Deep Sort Algorithm

Today, autonomous cars can drive on their own and detect any objects on the road to drive safely.  Tesla autonomous cars are very popular nowadays, and other car companies GM, Lucid motors, etc. also in the same technology.

In another 5 years, you can see more autonomous cars on the road, and you can see 80% to 90% of electric autonomous cars on the road by the end of 2050.

What is Yolov5 DeepSort Pytorch?

The detections generated by YOLOv5, a family of object detection architectures and models pretrained on the COCO dataset, are passed to a Deep Sort algorithm which tracks the objects. It can track any object that your Yolov5 model was trained to detect.

Source: Mikel Brostorm

In this blog, we will see the standard code and use our own video to check how it works.

What

You can find a code on my Google Colab and Yolov5 + Deep Sort with PyTorch.

Copy to Clipboard
Copy to Clipboard
Copy to Clipboard
Copy to Clipboard
Copy to Clipboard
Copy to Clipboard

I was trying to check my own video, however, it was throwing the error “Format mov,mp4,m4a,3gp,3g2,mj2 detected only with low score of 1, misdetection possible!”.

Issue number 121 created on this open-source library and waiting for the developer’s reply.

I will update the post once it has been fixed and get the solution to work on our own video files.

Update on 14th July:

The developer reported that the download file option only works in AVI file format, not on mp4 format.  I have updated the core now.

By |2021-07-14T22:29:33+00:00July 12th, 2021|Artificial Intelligence|0 Comments

About the Author:

Leave A Comment