I will build a custom computer vision model for object detection or classification
About this Gig
Want software that can see or hear? I build custom computer vision and audio models trained on your own data, using whatever architecture fits your problem best.
What I build:
- Object detection: count, locate and track objects (YOLO, Faster R-CNN, DETR, SSD or any other model)
- Image classification and custom CNNs (defects, products, plants, documents, medical images)
- Segmentation and custom visual models for your specific use case
- Audio classification models
- Small, fast models for phones, Raspberry Pi and other edge devices (ONNX, TFLite, quantization)
What you get:
- A trained model tested on your data, with an honest accuracy report
- Clean Python code (PyTorch or TensorFlow) and a simple script or app to run it
- Advice on collecting and labelling data if you don't have it yet
I have an M.Sc. in Data Science and specialise in deep learning and computer vision.
Message me BEFORE ordering with example images and what the model should detect. I'll tell you what's realistic and which package fits.
Programming language:
Python
•
Colab
Tools:
Jupyter Notebook
•
OpenCV
•
TensorFlow
•
CVAT
•
Colab
Frameworks:
Scikit-learn
•
Keras
•
PyTorch
Other Data Science & ML Services I Offer
FAQ
Do I need labelled data?
Not always. For simple tasks a pretrained model is enough; otherwise I will help you label a small dataset.
Can it run on a small device?
Yes. The Premium package optimises the model for edge devices like a Raspberry Pi or a phone.
