Hey, @suresh.krishna @arvindchandna,
I hope you both are doing well.
My name is Krapanshu Tyagi, and I am currently a third-year B.Tech student in Computer Science and Engineering at VIT University, India. I am very interested in contributing to the GSoC 2026.
My primary area of interest is AL/ML, particularly deep learning and explainable AI. I have previously worked with my professors on implementing research papers in the healthcare AI domain, including work on explainable models for heart stroke prediction and early Alzheimer’s detection. I am currently working on a paper on quantum-inspired energy-efficient AI models for ECG-based cardiac disease classification. Through these projects, I’ve gained experience working with model training pipelines, medical datasets, evaluation metrics, and experimental validation.
I also have experience in web and app development, which I believe could be useful for system integration aspects of this project.
I am currently setting up the automated-preferential-looking repo locally to better understand the existing pipeline. But for now I have few questions:
- For GSOC 2026, is the main emphasis excepted to be on improving the model or more on system integration?
- Also, is there any specific dataset, technology or research paper you would recommend exploring early on, to prepare for this project?
I am super excited to start contributing early and would really appreciate your guidance on where I should focus first.
Thank you for your time.
Thank you for the interest. We will not be offering this project in GSoC 2026. All the best with your search for a good project.
Suresh
Thank you for letting me know.
I thought it was offered as it was mentioned in the project ideas list for GSoC 2026.
Hello Sir,
I’m Sanhitha Reddy, a Computer Science undergraduate with a strong focus on Python, deep learning using PyTorch, and computer vision. I’m very interested in the project on developing a deep-learning-based eye-tracking system for measuring visual function in infants and young children.
The clinical relevance of objectively assessing visual development at an early age, combined with the technical challenge of building a reliable gaze-tracking pipeline from unconstrained infant video data, makes this project particularly exciting to me. I’m especially interested in how the tracking module, visual stimulus presentation, and analysis pipeline come together to form a complete and usable proof-of-concept system.
Through my recent work, I’ve been building end-to-end deep learning workflows for image-based data, with a strong emphasis on structured experimentation, modular implementation, and reproducible evaluation. Because of this, I’m keen to contribute to improving the robustness and evaluation of the eye-tracking model, as well as supporting the integration of the different components into a unified application.
I’ve started going through the repository and would love guidance on where it would be most useful to begin — whether that is understanding the current training pipeline, experimenting with the existing models, or working on specific open issues.
GitHub: sanhithaac (Allampati Chenchu Sanhitha Reddy ) · GitHub
LinkedIn: https://www.linkedin.com/in/allampati-chenchu-sanhitha-reddy-47a438352/
Looking forward to contributing and learning from the community.