Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a real time object detection and tracking system
Ahmad-lou
About this gig
Need a system that detects and tracks objects in real-time from video or camera feeds?
I build production-ready computer vision pipelines using YOLO and DeepSort fast, accurate,and built for real-world conditions.
I've worked on serious CV projects including:
- Real-time multi-object safety monitoring systems (YOLO + DeepSort)
- Custom YOLO models fine-tuned on domain-specific datasets (95%+ accuracy)
- Zone-based monitoring and hazard detection logic
- Full pipelines from raw video detection tracking API output
WHAT I CAN BUILD FOR YOU:
- Real-time object detection from video files or live camera streams
- Multi-object tracking with persistent IDs across frames (DeepSort)
- Custom YOLO model training on your own dataset
- REST API endpoint that receives images/video and returns predictions
- Dataset labeling pipeline (Roboflow) and training from scratch
TECH STACK:
Python | YOLOv5 / YOLOv7 / YOLOv8 | DeepSort | OpenCV | FastAPI | Roboflow
PROVEN RESULTS:
- Multi-child real-time safety monitoring live tracking + hazard logic Asian card game
- object detection 13 classes, 95%+ precision/recall, ordered spatial output, deployed as API
Get to know Ahmad-lou
Ahmad-lou
Python Developer And AI Engineer
- FromMorocco
- Member sinceJun 2026
- Avg. response time1 hour
Languages
English, Arabic, French
Hi! I'm Hamid, a Python developer and AI engineer with 4+ years of hands-on experience building
intelligent systems that actually work.
I specialize in:
✅ AI chatbots & agents
✅ RAG systems for document Q&A
✅ Workflow automation (n8n)
✅ NLP models & text analysis
✅ Machine learning & data science
I hold a Master's in Business Intelligence & Analytics and have delivered 20+ projects on Upwork. I write clean code, communicate clearly, and always deliver on time.
Whether you need a smart chatbot, a data pipeline, or a custom automation — I'm here to make it happen.
Let's talk!
My Portfolio
Other AI Development Services I Offer
FAQ
I don't have a labeled dataset — can you still help?
Absolutely! I can guide you through the full dataset preparation process using Roboflow — from raw images to fully labeled, training-ready data. This can be added as an extra to any package.
Will the system work on a live camera or only on video files?
Both! I can build the pipeline to work with pre-recorded video files, live webcam feeds, or RTSP streams from IP cameras. Just mention your setup when ordering and I'll configure it accordingly.
How accurate will the detection be on my specific use case?
Accuracy depends on your data quality and object complexity. With a clean, well-labeled dataset, YOLO models typically reach 90–95%+ precision. For best results, I recommend at least 200–500 images per class. I always share evaluation metrics (precision, recall, mAP) before final delivery.
Can I integrate the system into my existing app or website?
Yes. The Premium package includes a FastAPI REST endpoint that receives images or video frames and returns predictions in JSON format — ready to plug into any app, website, or backend system regardless of the tech stack you use.
