Class Incremental Learning for Image Segmentation
Budget: ₹1,500 – ₹12,500 INR
We are looking for a skilled developer to work on a project involving class incremental learning for image segmentation by applying knowledge distillation. The ideal candidate should have experience in the following:
Programming Language:
- Proficiency in Python is preferred for this project.
Image Segmentation Algorithm:
- Familiarity with the U-net algorithm is required for this project.
- Familiarity with the Incremental Learning algorithms
- Focusing on pixel segmentation for each class, not only classification.
Model Preference:
- The client has no specific preference for a pre-trained model or a model built from scratch.
Skills and Experience:
- Strong programming skills in Python.
- Experience with image segmentation algorithms, specifically U-net or models that are suitable for my dataset.
- Familiarity with class incremental learning techniques.
- Ability to work with pre-trained models or build models from scratch.
Project deliverables:
- Develop and implement a class incremental learning approach for image segmentation using the U-net or CIL algorithms.
- Ensure compatibility with the preferred programming language, Python.
- Provide documentation and code comments for easy understanding and future maintenance.
Dataset:
My dataset is x-ray images, and the annotation is a JSON file for each class (coco format). I have three classes (permanent teeth as PERM, primary teeth as PRIM, and filling).
If you have the necessary skills and experience in Python programming, image segmentation algorithms, and class incremental learning, and are interested in working on this project, please submit your proposal.
Programming Language:
- Proficiency in Python is preferred for this project.
Image Segmentation Algorithm:
- Familiarity with the U-net algorithm is required for this project.
- Familiarity with the Incremental Learning algorithms
- Focusing on pixel segmentation for each class, not only classification.
Model Preference:
- The client has no specific preference for a pre-trained model or a model built from scratch.
Skills and Experience:
- Strong programming skills in Python.
- Experience with image segmentation algorithms, specifically U-net or models that are suitable for my dataset.
- Familiarity with class incremental learning techniques.
- Ability to work with pre-trained models or build models from scratch.
Project deliverables:
- Develop and implement a class incremental learning approach for image segmentation using the U-net or CIL algorithms.
- Ensure compatibility with the preferred programming language, Python.
- Provide documentation and code comments for easy understanding and future maintenance.
Dataset:
My dataset is x-ray images, and the annotation is a JSON file for each class (coco format). I have three classes (permanent teeth as PERM, primary teeth as PRIM, and filling).
If you have the necessary skills and experience in Python programming, image segmentation algorithms, and class incremental learning, and are interested in working on this project, please submit your proposal.