Video Annotation with SAM 2 + YOLO Fine-Tuning

In this tutorial, we walk through a complete real-world computer vision pipeline — from annotating a hospital room video dataset to fine-tuning a custom YOLO model for multi-object detection and segmentation.

What you'll learn:
→ How to annotate video datasets using Labellerr with AI-assisted tools
→ Using SAM 2 (Segment Anything Model 2) to auto-track objects across video frames
→ Exporting annotations in JSON format for model training
→ Converting annotations to YOLO training format using convert_from_manifest
→ Fine-tuning YOLOv8 segmentation model on a custom dataset (100 epochs, 500 images)
→ Running inference on images and videos to detect hospital room components

Use Case:
We trained a model to detect real medical equipment such as mattresses, humidifiers, infusion setups, vital signs monitors, and more — all from raw, unannotated video data.

GitHub: https://github.com/Labellerr/Hands-On-Learning-in-Computer-Vision
Notebook: https://github.com/Labellerr/Hands-On-Learning-in-Computer-Vision/blob/main/fine-tune%20YOLO%20for%20various%20use%20cases/Components_Recognition_in_Hospital_Rooms.ipynb
#ComputerVision #YOLOv8 #ObjectDetection #SAM2 #SegmentAnything #VideoAnnotation #Labellerr #MachineLearning #DeepLearning #Ultralytics #ModelTraining #AITutorial #ArtificialIntelligence #ImageSegmentation #CustomModel #YOLOTraining #VideoDataset #DataAnnotation #MLOps #PythonAI #MedicalAI #HealthcareAI #RealTimeDetection #FineTuning #ComputerVisionTutorial #AIEngineer #ObjectSegmentation #InstanceSegmentation #OpenCV #AIProjects

Видео Video Annotation with SAM 2 + YOLO Fine-Tuning канала Labellerr AI