Object detection using YOLO (python) ([High School Information Department Information II] Teacher training materials)

Introduction

As the name implies, YOLO (You Only Look Once) is an epoch-making algorithm that recognizes and detects an object just by looking at it once like a human being. Compared to the conventional method, it has the features of high speed processing, strong distinction between object and background recognition, and easy generalization. This time, we will use YOLO and python to detect objects on the photograph where objects are detected using Tiny YOLO and R in the teaching materials.

Teaching materials

[High School Information Department "Information II" Teacher Training Materials (Main Volume): Ministry of Education, Culture, Sports, Science and Technology](https://www.mext.go.jp/a_menu/shotou/zyouhou/detail/mext_00742.html "High School Information Department "Information II" teaching materials for teacher training (main part): Ministry of Education, Culture, Sports, Science and Technology ") Chapter 3 Information and Data Science Second Half (PDF: 7.6MB)")

environment

• ipython
• Colaboratory - Google Colab

Parts to be taken up in the teaching materials

Learning 18 Text mining and image recognition: "3. Object detection using TinyYOLO"

Implementation example and result in python

This time, clone (copy) darknet from the repository of github and perform image recognition using the trained weight data yolov3.weights of YOLOv3. This time it is written as an implementation example in python, but in order to run YOLOv3 on darknet, I would like to focus on command execution so that I can detect objects in a way that I do not code by myself as much as possible.

!git clone https://github.com/pjreddie/darknet

The execution result is as follows

Cloning into 'darknet'...
remote: Enumerating objects: 5913, done.
remote: Total 5913 (delta 0), reused 0 (delta 0), pack-reused 5913
Receiving objects: 100% (5913/5913), 6.34 MiB | 9.93 MiB/s, done.
Resolving deltas: 100% (3918/3918), done.

Now that you have a git clone, move to the darknet directory and run make.

import os

os.chdir('darknet')
!make

After the make is completed, download the trained weight data yolov3.weights of YOLOv3 to the same directory. This time I used the wget command.

!wget https://pjreddie.com/media/files/yolov3.weights

I would like to have YOLO detect the object in the main subject, but this time I would like to use giraffe.jpg in the data directory to detect the object.

!./darknet detect cfg/yolov3.cfg yolov3.weights data/giraffe.jpg

layer     filters    size              input                output
    0 conv     32  3 x 3 / 1   608 x 608 x   3   ->   608 x 608 x  32  0.639 BFLOPs
    1 conv     64  3 x 3 / 2   608 x 608 x  32   ->   304 x 304 x  64  3.407 BFLOPs
    2 conv     32  1 x 1 / 1   304 x 304 x  64   ->   304 x 304 x  32  0.379 BFLOPs
    3 conv     64  3 x 3 / 1   304 x 304 x  32   ->   304 x 304 x  64  3.407 BFLOPs
：
  103 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs
  104 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs
  105 conv    255  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 255  0.754 BFLOPs
  106 yolo
Loading weights from yolov3.weights...Done!
data/giraffe.jpg: Predicted in 19.677707 seconds.
giraffe: 98%
zebra: 98%

Giraffes and zebras were detected. Let's take a look at the actual detected image.

from IPython.display import Image

Image("predictions.jpg ")

I was able to detect it successfully.

Source code

https://gist.github.com/ereyester/46a25e70c866c581320a66a77153aa2d