[gsoc-2026] SwiftGestureKit: Dual-Hand Vision Recognition & Documentation Proposal

Development GSoC
1

Hello Swift community,

I am applying for Google Summer of Code 2026 with the Swift project and would really appreciate feedback on my proposal: SwiftGestureKit — Dual-Hand Vision Recognition & Swift Developer Experience Improvements.

While building HastaAkshar, a real-time Bharatanatyam mudra recognition app using the Vision framework, I encountered an underdocumented behavior in VNDetectHumanHandPoseRequest: the default maximumHandCount effectively results in only one reliable observation being processed, which caused false positives for double-hand gestures. Debugging this took significant time, and I could not find any canonical Swift resource explaining the issue.

I also encountered UIKit migration confusion around UIButton.Configuration in iOS 15+, where direct color assignments are silently overridden.

Based on these experiences, my proposal focuses on:

• SwiftGestureKit — A Swift package providing a structured dual-hand Vision pipeline, confidence-weighted gesture scoring, and Swift Concurrency support
• DocC-based documentation and tutorials demonstrating correct multi-hand gesture handling
• Example-driven developer guidance addressing real-world pitfalls discovered during app development

I would really appreciate feedback on:

  • Whether this scope is appropriate for Swift GSoC

  • Possible alignment with Swift documentation tooling (DocC, examples, tutorials)

  • Suggested repositories where parts of this work could integrate with existing Swift efforts

  • Any recommendations on narrowing or refining the proposal

Thank you for your time and guidance.

1 Like
2

Thanks for the thread and idea.

I'm sorry but this isn't the kind of project we're going to accept in the Swift project. Projects in Swift project's GSoC are focused mostly around the language, tools and ecosystem in the sense of "more platforms support". We don't really run projects which are app focused libraries like this.

This does sound like a nice project by itself, however it is not a good proposal idea for Swift's GSoC.

Hope this helps

2 Likes