Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a computer vision and object detection system
Hammad Durrani
About this gig
Need a system that can see, analyze and understand
images and video in real time?
I build custom computer vision systems for object
detection, image classification, face recognition,
OCR, pose estimation and video analytics trained
on your specific data and deployed production-ready.
What I deliver:
· Object detection using YOLOv8, YOLOv11 and
custom trained models
· Image classification with CNN and ResNet
· Face detection and recognition systems
· OCR for document and text extraction
· Pose estimation and human activity detection
· Real-time video analytics and processing
· Custom model training on your dataset
· REST API for image and video inference
· React or Next.js frontend with live preview
· Deployment on AWS, GCP or Hostinger
· Full source code and documentation included
Perfect for:
· Security and surveillance systems
· Quality control in manufacturing
· Medical imaging and diagnostics
· Retail analytics and people counting
· Document scanning and OCR automation
· Sports and fitness tracking apps
Message me before ordering share your use case
and sample images and I will confirm exactly what
I can build for you.
Get to know Hammad Durrani
Hammad Durrani
AI Full Stack Engineer and Custom Software Builder
- FromPakistan
- Member sinceMar 2026
Languages
Urdu, Pashto, English
AI Full Stack Engineer & Custom Software Builder
— custom ML models, LLMs, RAG systems, agentic pipelines, autonomous AI agents, computer vision ,NLP, and full stack custom software development. From fine-tuning LLMs to building intelligent
agents, predictive models, document intelligence, AI chatbots, SaaS products, and complete business
automation — everything delivered production-ready with clean code, full documentation, and live
deployment on AWS, GCP, Vercel, and Hostinger with custom domain and SSL included.
From idea to live product — I build it all.
Message me before ordering.
My Portfolio
FAQ
What image formats do you support?
I support JPG, PNG, MP4, AVI and live camera feeds. I can also process video streams in real time depending on your requirements.
Do I need to provide training data?
For custom detection you need labeled images of your objects. Minimum 100 images per class recommended. I can help with data collection and annotation if needed.
How accurate will the model be?
Accuracy depends on data quality and quantity. With good labeled data I typically achieve 85 to 95 percent accuracy. I share evaluation metrics with every delivery.
Can it work in real time on video?
Yes. Premium package includes real time video processing optimized for speed using YOLO and OpenCV with GPU acceleration support.
Will I get the trained model file?
Yes. You receive the complete trained model weights, source code, training scripts and documentation with every delivery.
