Browse categories
Explore
Fiverr Pro
English
$
USD
I will train custom yolov8 object detection models using pytorch
Younes Belhadj
About this gig
Are you looking for a custom Computer Vision solution or a trained object detection model tailored to your specific dataset?
I build, fine-tune, and deploy high-performance computer vision pipelines using PyTorch, YOLOv8, ResNet, and OpenCV. Whether you need real-time object tracking, defect detection, or automated visual analytics, I deliver fast and reliable end-to-end models built for production.
Services Offered: Custom Object Detection: Training and fine-tuning YOLOv8/ResNet models on custom image/video datasets. Dataset Preparation & Annotation: Converting and formatting datasets to YOLO, COCO, or Pascal VOC formats with image augmentation. Performance Optimization: Hyperparameter tuning, precision evaluation (mAP, Precision/Recall), and model quantization/export (ONNX, TensorRT). Real-Time Video Analytics: Custom OpenCV pipelines for live stream tracking, counting, and automated alert logging. Production Deployment: Clean Python integration code, API endpoints, or single-click Docker containers.
Tech Stack: PyTorch, YOLOv8, ResNet, OpenCV, CUDA, Python, Docker.
Whether you need industrial quality control (weld defect detection), security monitoring, traffic analysis, or reta
Get to know Younes Belhadj
Younes Belhadj
Data Science and AI Engineer
- FromAlgeria
- Member sinceMay 2020
- Avg. response time1 hour
Languages
English, Arabic, French
Hello! I’m a Data Science & AI Engineer specializing in Computer Vision (YOLOv8, ResNet), RAG systems, and full-stack web platforms (React, Node.js, Express, Docker).
I build intelligent end-to-end software—from custom AI/ML model deployment and private document chatbots to production-ready web apps and SaaS tools.
What I bring to your project:
• Tailored AI/ML model architecture & integration
• Scalable full-stack web apps with modern UI
• Containerized code ready for deployment
• Fast communication & reliable delivery
Let's discuss your project!
FAQ
What formats do I need to provide my dataset in?
I can work with raw images/videos or pre-annotated datasets in YOLO, COCO, Pascal VOC, or JSON formats. If your data isn't annotated yet, I can assist with labeling and formatting it for training.
Can you train models on private or non-public datasets?
Yes, absolutely. Your data remains 100% confidential and is processed locally on isolated hardware. None of your data or custom-trained weights will ever be shared or reused.
Which YOLO architecture version do you use?
I primarily work with YOLOv8 and YOLOv11 for high accuracy and speed. However, I can also train, fine-tune, or convert legacy models (YOLOv5, YOLOv7) depending on your target system requirements.
What if my model does not achieve the desired accuracy?
I evaluate models using standard metrics (mAP, Precision, Recall). If accuracy is low, I apply data augmentation, hyperparameter tuning, or class balancing to optimize performance before delivery.
