Real-Time Sign Language Recognition
A system that recognises a working vocabulary of sign language from a webcam and renders it as text in real time, built on pose landmarks rather than raw video. Choose a regional sign language and be honest that continuous signing is a research problem, not a semester one.
Advanced
Needs a team, real planning, and a supervisor who knows the area.1 semester, 2 to 3 students
5 to ship
3 optional extrasSuggested stack
What you should ship
- Recorded dataset covering at least 50 signs, multiple signers, varied lighting and backgrounds
- Landmark extraction pipeline producing hand, face and upper body keypoints per frame
- Sequence classification model with accuracy reported on signers held out of training
- Live demonstration application running at a usable frame rate on a normal laptop
- Confusion analysis identifying which signs are mistaken for which, and why
If you have time left
- Short phrase recognition rather than isolated signs
- A guided practice mode that tells a learner which part of their sign was wrong
- Browser deployment via an exported model running client side
The problem
Sign language users have very little software that meets them where they are. Most accessibility tooling assumes speech or text, and the gap is widest exactly where it matters, in unplanned everyday interactions.
What you build
A pipeline that extracts pose landmarks from webcam frames, a model that classifies the resulting sequence into a sign, and an application that displays recognised signs live.
Why landmarks rather than raw video
Working from extracted keypoints rather than pixels reduces the data requirement enormously, trains on hardware you actually have, and generalises better across lighting and clothing. It is the difference between a project that works and one that runs out of time.
The hard part
Signer independence. A model trained on three people will score well on those three and fail on a fourth, and this is the single most common flaw in student projects on this topic. Hold at least one signer out of training entirely and report that number as your headline result.
How to be honest about scope
Continuous signing, with grammar, facial grammar markers and coarticulation, is an open research problem. A fixed vocabulary of isolated signs is an achievable and worthwhile project. Say which one you are doing in your first sentence, and do not let the report imply otherwise.
Ethics
Involve deaf signers in the design if you possibly can. Projects in this area built entirely by hearing students frequently solve a problem the community does not have.
Ideas and guidance, not finished projects
These are project ideas and scoping guidance, published free for students to use as a starting point. I do not build, write, or sell final-year projects, and I do not complete coursework for anyone. Take an idea, make it yours, and build it.
Not sure this one fits you?
I help students pick an idea that matches their skills and their deadline, then scope it down until it is finishable. Guidance only, never done for you.
Stuck on something specific?
Writing only gets you so far. If you want an answer to your situation rather than the general case, book a session and we will work through it together. Every session is free; a few slots open each week.
Follow along
New writing, resources and project ideas land here first.