Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a custom image segmentation model with pytorch
Saniya Khan
About this gig
I've built two segmentation systems that are live right now:
Brain tumor detection on MRI scans Attention UNet, 76.6% IoU improvement over standard UNet
Flood area detection from drone imagery deployed for disaster response use cases
Both are on HuggingFace as real interactive apps, not just notebooks.
What I'll deliver:
- Custom UNet or Attention UNet trained on your dataset
- Grad-CAM heatmaps so you can see what the model is focusing on
- Streamlit web app your team can use without touching code
- Clean, documented source code you fully own
I work with medical imaging, satellite imagery, industrial inspection, and general CV datasets.
Not sure if segmentation is right for your problem? Message me first I'll give you an honest answer.
Tech: PyTorch · OpenCV · Streamlit · HuggingFace · Python
Get to know Saniya Khan
Saniya Khan
Computer Vision and AI Engineer
- FromPakistan
- Member sinceFeb 2024
- Avg. response time9 hours
Languages
English
I specialize in Computer Vision and Deep Learning using PyTorch. I build image segmentation systems, object detection models, and AI-powered web apps.
✅ Brain Tumor Segmentation (UNet, Attention UNet, Grad-CAM)
✅ Flood Area Segmentation (Drone Imagery)
✅ Streamlit Web Apps
I deliver clean, well-documented code with fast turnaround. Let's build something amazing!
FAQ
What type of images can you segment?
I can work with medical images (MRI, CT), satellite imagery, drone footage, and general images. Message me with your dataset details!
Do you provide the source code?
Yes! All packages include full Python source code with documentation.
What do you need from me to start?
I need your dataset (images + masks), the number of classes, and any specific requirements.
