In this video, you will get to know the practical implementation of YOLO algorithm following all the concepts discussed in the previous video. For labeling custom dataset images, LabelImg tool has been used. Instructor used Google Colab for training model on custom dataset and for testing purposes, we need to install openCv library. The model implementation discussed in this video includes;
– Testing model on pre-trained famous COCO dataset which performs the object detection task for up-to 80 classes
– Training a model on custom dataset and then testing custom-trained model.

In this implementation part of the tutorial following topics will be covered:
00:00 Introduction
00:35 Custom Model Training
01:45 Creating Folder on Drive
02:32 Google Colab Notebook
06:20 Cloning Darknet Repository
15:18 Model Compilation
16:46 Generate Training and Testing Files
19:53 Pre-Trained Weights
23:04 Model Ready for Training
23:37 Start Custom Model Training
25:38 Finished Custom Model Training
26:11 Custom Model Testing
46:09 Testing Result of Custom Model
47:00 Pre-Trained Model Testing
47:05 Necessary Files for Pre-Trained Model Training
47:45 Implementation of Pre-Trained Model Testing
51:45 End of Implementation

Researchers: Ms. Nayab Tahir, Ms. Noshiza Mukhtar, Ms. Khadija Akhter, Ms. Simran Parveen

Important Resources
Custom datasets download link:
https://public.roboflow.com/object-detection
Darknet Repository:
https://github.com/AlexeyAB/darknet.git
Custom weights:
http://pjreddie.com/media/files/darknet53.conv.74
Train and test data generator: https://github.com/jakkcoder/training_yolo_custom_object_detection_files
Coco classes file
https://github.com/matlab-deep-learning/Object-Detection-Using-Pretrained-YOLO-v2/blob/main/+helper/coco-classes.txt
Coco yolov3.cfg file
https://github.com/pjreddie/darknet/blob/master/cfg/yolov3.cfg

please visit the following link to understand the theoretical background:

For dataset and notebook please visit:
https://drive.google.com/drive/folders/1B9ZLAooHzykeWXD6ARRjms2-24qzTdvx?usp=sharing

#tazzainamalik #yoloalgorithm #yoloobjectdetectionalgorithm #objectdetectiondeeplearning
#objectdetection

source