Interactive Web Tool for Satellite Image Segmentation
Budget: €1,500 – €3,000 EUR
I'm looking for a developer to create a web-based tool that facilitates scribble-based weakly-supervised semantic segmentation on satellite images. This interactive interface should be simple and minimalistic, allowing users to draw scribbles on images and categorizing similar pixels effortlessly.
Key Features:
- Real-time fine-tuning of a lightweight, pre-trained neural network for semantic segmentation as users annotate the images.
- Capability to segment various classes including buildings, roads, vegetation, and small objects.
- A straightforward, user-friendly interface devoid of unnecessary complexity.
Ideal Skills and Experience:
- Proficiency in web development with an emphasis on creating interactive interfaces.
- Experience in machine learning, particularly with neural networks and semantic segmentation.
- Understanding of satellite images and their specific characteristics (TIFF, JPG, WEBP).
- Ability to design simple, minimalistic user interfaces without compromising functionality.
Key Features:
- Real-time fine-tuning of a lightweight, pre-trained neural network for semantic segmentation as users annotate the images.
- Capability to segment various classes including buildings, roads, vegetation, and small objects.
- A straightforward, user-friendly interface devoid of unnecessary complexity.
Ideal Skills and Experience:
- Proficiency in web development with an emphasis on creating interactive interfaces.
- Experience in machine learning, particularly with neural networks and semantic segmentation.
- Understanding of satellite images and their specific characteristics (TIFF, JPG, WEBP).
- Ability to design simple, minimalistic user interfaces without compromising functionality.