GSoC 2026 Guidelines for Contributions and Proposals

[henrykironde]

Contribution and Proposal Guidelines

🤖 AI Usage

• Please review the AI guidelines on the Ideas page before submitting contributions or proposals.
• If you use AI tools or agents while developing your proposal or contributions, clearly cite them.

---

🚫 What Not to Do

• Do not repeatedly request PR reviews. Opening a PR is sufficient to notify maintainers that a review is needed. Maintainers may be busy, so please be patient.
• Do not send emails.
• Do not submit CVs.
• Do not open new discussions if an active discussion on the same topic already exists. Always check existing issues and discussions before opening a new one to avoid duplication.

---

🔀 Pull Requests (PRs)

• All PRs are evaluated equally, whether they are merged or not.

• Evaluation considers several factors including:

  • The issue being addressed
  • The quality of the solution
  • Design decisions
  • Overall contribution to the project

• Typo and link fixes are appreciated, but they do not significantly impact evaluation since the primary focus is on problem solving and design contributions.

---

📖 Familiarize Yourself with the Codebase

Before contributing, please review the documentation:

• Official documentation:
• Developer documentation

---

🛠️ Contribute Before Proposing

We strongly encourage contributors to demonstrate familiarity with the project by solving some Good First Issues

This helps you better understand the codebase and development workflow.

---

✍️ Writing Your Proposal

When creating your proposal:

• Clearly describe the problem you are addressing.
• Provide a well-structured solution or approach.
• Explain your design decisions and expected impact.

---

🔧 Develop Your Prototype

Continue refining and developing your proposed solution. Iteration and improvement are expected as part of the process.

---

✅ Evaluation and Selection

Proposals will be reviewed based on:

• Quality of the idea and implementation
• Technical feasibility
• Design and problem-solving approach
• Demonstrated contributions to the project

---

[vickysharma-prog]

Hi @henrykironde,

Thanks for sharing these guidelines!

I've been actively contributing to DeepForest and have started exploring an approach for recovering bird annotations from historical imagery. I've started building a prototype using a hybrid approach combining dot detection from screenshots with segmentation-based bird detection, then aligning both to produce DeepForest-compatible annotations. I also opened discussion #1333 to better understand the technical direction and request a few sample images for testing.

As I continue refining the prototype and drafting my proposal, what would be the best way to get feedback from mentors to ensure the approach aligns with the project goals?
Thanks!

[musaqlain]

hello mentors, can u review my proposal, and give me some feedback?

Proposal 2: Recovering historical image data using automated ortho-registration and image-matching

I’ve given commenter access to this document to you via your respective emails. (not notified any)
thanks 🙌
cc @bw4sz @jveitchmichaelis