Door categorieën bladeren
Ontdekken
Fiverr Pro
Nederlands
$
USD
I will build computer vision models using yolo opencv and CNN
Machine Learning, NLP Specialit
Ranked 200 th among 3000+ Expert
competitors in the Trace the Ace NLP competition.
My work focuses on end-to-end Machine Learning and NLP from project architecture, EDA and feature engineering to mo...
Over deze dienst
Need a custom computer vision model for your images or videos?
I build computer vision solutions using YOLO, OpenCV, CNN, TensorFlow and PyTorch, tailored to your dataset and goals.
Portfolio:https://kawsar-portfolio-kawsar07ahmmed0712-5454.vercel.app/
My services:
- Object detection and tracking with YOLO
- Image classification using CNN and transfer learning
- Image processing with OpenCV
- Image segmentation on suitable labeled datasets
- Deep learning model training, fine-tuning and evaluation
- Flask/FastAPI inference APIs and optional web demos
What you receive (based on your package):
- Dataset review, preprocessing and augmentation
- Trained model and tested inference script
- Clean Python source code and documentation
- Evaluation metrics (accuracy, F1, mAP or IoU as appropriate)
Packages:
- Basic: One small, labeled classification or detection task.
- Standard: Custom training, optimization and evaluation.
- Premium: End-to-end model pipeline with a local inference API.
Annotation and cloud hosting are not included. No fixed accuracy is guaranteed before dataset evaluation.
Message me with your task, dataset size, classes and target environment before ordering.
_uchn0e.jpg)
Expertise:
Afbeeldingenverwerking
•
Classificatie
•
Objectherkenning
Programmeertaal:
Python
Tools:
Jupyter-notitieboek
•
opencv
•
tensorflow
•
Colab
•
PyTorch
Frameworks:
keras
•
PyTorch
Mijn portfolio
Andere Data science en ML diensten die ik aanbied
Veelgestelde vragen
What computer vision services do you offer?
I develop custom computer vision solutions for YOLO object detection, CNN image classification, OpenCV image processing, and suitable image segmentation or tracking tasks. The approach depends on your dataset, project goals, and selected package.
Can you train a model using my own dataset?
Yes. I can train or fine-tune a model using your images or video dataset. Please provide sample files, dataset size, class information, and existing labels. I will review the requirements and recommend an appropriate training approach.
What if my dataset is not labeled or annotated?
Data annotation is not included by default. Classification requires class labels, while object detection and segmentation generally require bounding boxes or masks. If your dataset is unlabeled, please contact me for a separate annotation quote or preparation plan.
Which package should I choose?
Basic ($90) is for one small, labeled classification or detection task. Standard ($180) covers custom model training, testing, and optimization. Premium ($350) includes a more complete pipeline with a local inference API. Larger projects require a custom offer.
Will I receive the trained model and source code?
Yes. Deliverables include the trained model file or checkpoint, Python source code, an inference script, evaluation results, and documentation according to your package. You can run the model in a compatible environment following the provided instructions.
Can you guarantee 95% or 99% model accuracy?
No specific accuracy can be guaranteed before evaluating your dataset. Performance depends on data quality, quantity, labeling, class balance, and task complexity. I provide appropriate evaluation metrics such as accuracy, precision, recall, F1-score, mAP, or IoU.
Can you build a Flask or FastAPI application for my model?
Yes. Premium includes a basic local inference API using Flask or FastAPI. A simple web interface for uploading images and viewing predictions can be added as a Gig Extra. Integration with existing applications may require a custom offer.
Do you support real-time detection and cloud deployment?
Real-time detection can be assessed based on your target hardware, video resolution, model size, and required FPS. Cloud deployment, edge-device optimization, and hosting are not included by default. Please contact me for feasibility assessment and custom pricing.
What is included in a revision?
Basic includes 1 revision, Standard includes 2, and Premium includes 3. Revisions cover corrections within the originally agreed project scope. New datasets, additional classes, different architectures, or new features may require an extra charge.
What information should I provide before ordering?
Please share your project goal, sample images or videos, dataset size, number of classes, annotation status, preferred framework (if any), target hardware or platform, expected output, and deadline. I recommend discussing complex requirements before placing an order.
