FYP question

[lamelo03]

Search before asking

• [X] I have searched the HUB issues and discussions and found no similar questions.

Question

Hi there currently im working on my final year project for my College. this is the title: Machine Learning to detect Video Anomaly between repeated test runs. Oppstar Technology (1).xlsx

This project aims to develop an automated testing and anomaly detection system for repeated test executions in a PC environment. Using a KVM setup, the project will establish a repeatable test sequence that emulates keystrokes and mouse actions to simulate user behaviour. Each test run will be captured by a video capture device, and a library of test videos will be stored for analysis. Through image processing and machine learning techniques, a baseline database will be created to record key parameters observed during the tests. This baseline will serve as a reference point to identify any deviations or anomalies in subsequent test runs, thus allowing for real-time detection of abnormal behaviour in the test environment. There are two machine learning used in this project, the first one is YOLO for error detection like error code popping out from the PC, another one is OCR for character recognition. The outcome of this project is expected to enhance automated testing reliability by leveraging AI-driven image analysis to ensure consistent test execution and accurate anomaly detection.

this is the projects, which YOLO version should i use and where do i start

Additional

No response

[UltralyticsAssistant]

👋 Hello @lamelo03, thank you for reaching out about your final year project using the Ultralytics HUB 🚀! This is an automated response, but rest assured, an Ultralytics engineer will assist you soon.

For a project like yours—automated testing and anomaly detection across video recordings—our HUB documentation is a great place to start. Here are some resources that could help:

• Quickstart: Get started with training and deploying YOLO models swiftly.
• Models: Training and Exporting: Provides guidance on training YOLOv5 or YOLOv8 models, which would be used for your error detection tasks.
• Datasets: Preparing and Uploading: Instructions on formatting your dataset for training.
• Projects: Creating and Managing: How to organize your models and projects effectively.

Regarding which YOLO version to use, both YOLOv5 and YOLOv8 are powerful, but the choice may depend on factors such as your specific use case requirements and resource availability. YOLOv8 offers more recent advancements and improvements, so it might be a good starting point.

If you encounter any 🐛 issues while implementing your project, please provide screenshots and a minimum reproducible example, to help us assist you better.

We are committed to addressing all inquiries as quickly as possible. Thank you for your patience and good luck with your project! 😊

[pderrenger]

@lamelo03 hello! 😊 Your final year project sounds fascinating and ambitious, leveraging machine learning for anomaly detection in video captures. For your project, using YOLO for error detection is a great choice due to its efficiency and accuracy in real-time object detection tasks.

YOLO Version Recommendation

For your project, I recommend using YOLOv8, the latest version from Ultralytics. YOLOv8 offers improved performance and flexibility, making it well-suited for detecting error codes or anomalies in video frames.

Getting Started with YOLOv8

1. Ultralytics HUB: You can start by using the Ultralytics HUB to train your YOLOv8 model. It's user-friendly and allows you to upload datasets, train models, and deploy them seamlessly.

2. Dataset Preparation: Ensure your dataset is well-labeled with the types of errors or anomalies you want to detect. You can upload this dataset to the HUB for training.

3. Model Training: Use the pre-trained YOLOv8 model as a starting point. This will save you time and resources. You can customize the training settings to fit your specific needs.

4. Deployment: Once trained, you can deploy your model using the Ultralytics HUB App for real-time detection.

Additional Tools

For OCR (Optical Character Recognition), you might want to integrate a library like Tesseract or use a pre-trained OCR model that can work alongside YOLOv8 to recognize characters in your video frames.

Feel free to explore the Ultralytics HUB Docs for more detailed guidance on each step. If you encounter any issues or have further questions, don't hesitate to reach out here or join our Discord community for more interactive support.

Best of luck with your project! 🚀