Browse categories
Explore
Fiverr Pro
English
$
USD
I will train custom yolo models for object detection and computer vision
About this Gig
Need a custom computer vision solution for images or video? I train and deploy YOLO object detection models with Python, PyTorch and OpenCV, tailored to your dataset and use case.
What I Offer:
- Custom object detection and YOLO model training
- Image and video segmentation with SAM 2
- Object tracking, counting and trajectory analysis
- Dataset preparation, annotation and augmentation
- FastAPI integration and Docker deployment
- ONNX and TensorRT optimization for faster inference
- Mobile deployment with TFLite or CoreML inside your app
Use cases: sports video analysis, vehicle counting, product detection and defect inspection.
What You Get:
- Trained model weights and clean source code
- Evaluation metrics: mAP, precision, recall and FPS
- Sample predictions and a step-by-step setup guide
- API or deployment support, depending on your package
Accuracy and speed depend on your data, task and hardware, so we agree on the scope and success criteria before starting.
Send me your project details and a few sample images or videos, and I'll recommend the right package.
My Portfolio
FAQ
What do you need to get started?
A description of your use case, sample images or video footage, and your target performance requirements (such as real-time FPS or target accuracy). If you don't have labeled data, message me first.
Which YOLO models or architectures do you support?
I work with modern YOLO models (YOLOv8 through YOLO11 and newer) and SAM 2 for segmentation, and I choose the best one for your hardware (CPU, GPU or edge devices).
Can you help me integrate the trained model into my app or server?
Yes. The Premium package and extras include API development (FastAPI/Flask) and Docker deployment. I can also convert your model to TFLite or CoreML and add it to your mobile app.
How many images do I need?
For good results, usually 200–500 labeled images per class. With fewer images, I can use augmentation and pre-trained models. Send me samples and I'll advise.
My images aren't labeled. Can you help?
Yes. Add the Data Annotation extra and I'll label your images with bounding boxes.
Do I need a GPU or paid cloud server?
No. I handle training myself. For deployment, I optimize the model to run on your hardware, whether that's CPU, GPU, cloud or mobile.
