ABOUT EVENT

Workshop Description

This workshop focuses on the full workflow of training a YOLOv8 deep learning model for detecting Coconut Rhinoceros Beetles and building a simple Flask web application for running model inference. Attendees will learn how to prepare data, train and evaluate an object detection model, and deploy the model in a lightweight web interface. The workshop introduces practical skills in computer vision, Python-based machine learning, and basic web deployment. Attendees can expect to complete several hands-on activities during the workshop. They will set up the environment for YOLOv8 training, explore the CRB image dataset, and run a full training pipeline using Jupyter notebooks. Participants will also check model performance, try simple training adjustments, and export their trained model. They will then create a basic Flask web application that loads the trained YOLOv8 model, runs detection on uploaded images, and shows the results. They will learn how to organize project files, create the main Flask routes, and test the app on their own computer. By the end of the workshop, attendees will experience the full process from preparing a dataset to training a model to building a simple web demo.

Prerequisites

• Attendees should have a UH account
• BEFORE THE WORKSHOP participants should:

1. Have Python installed on their computer
2. Download the sample CRB image dataset
3. Download the example YOLOv8 and
4. Download the Flask scripts accessible from the shared Google Drive folder (linked below):https://drive.google.com/drive/folders/1eFJHFncNrIHWwPnAbVtkYgcmAmo5y58z

• Have the ability to run Jupyter notebooks on their computers.

Learning Objectives

This workshop aims to provide attendees with an understanding of how to prepare a dataset for YOLOv8, train an object detection model, and evaluate its performance. They will also learn how to load a trained model into a basic Flask web application, run inference on new images, and display detection results in a simple web interface. Participants will gain practical experience with the full workflow from model training to deployment.

Tools Used

• Jupyter notebooks
• Python
• Ultralytics YOLOv8 library
• Flask