Fingerspell — ASL Reader

Fingerspell reads the American Sign Language alphabet in real time, using nothing but the camera already in your phone.

Point the camera at your hand, form a letter, and hold it steady for a moment. The app tracks 21 points on your hand, recognises the letter, and builds it into a word. When you finish a word, it suggests corrections — and if the top suggestion isn't what you meant, you tap the one that is.

NO SPECIAL EQUIPMENT
There are no gloves, no sensors, no wearables, and nothing to plug in. If you have a phone with a camera, you have everything you need.

WORKS OFFLINE
Recognition runs entirely on your device. After the first launch there is no server, no upload, and no internet requirement. Your camera feed never leaves your phone.

FREE, WITH NO CATCHES
No account, no sign-up, no subscription, no adverts, no in-app purchases, and no data collection of any kind.

BUILT-IN ALPHABET GUIDE
New to fingerspelling? The app includes a reference chart and a written description of how to form each of the 26 letters, so you can learn as you go.

SMART WORD CORRECTION
Individual letters are sometimes misread — certain handshapes genuinely look alike to a camera. Rather than pretending otherwise, Fingerspell checks each finished word against a dictionary, weighted by how confident it was about each letter, and offers you the most likely matches to choose from.

WHAT THIS APP DOES NOT DO
Fingerspell recognises the fingerspelling alphabet. It does not translate full American Sign Language, which has its own grammar, facial expressions, and continuous signing that go well beyond letters. Fingerspelling is one real and widely used part of ASL — commonly used for names, places, and words without a dedicated sign — and that is what this app is for.

Two letters, J and Z, involve movement rather than a fixed handshape. The app approximates them from their starting position, so they are less reliable than the other 24 letters.

ABOUT
Fingerspell was built as an MSc research project at Atlantic Technological University, Ireland. The goal was to find out how useful an accessibility tool can be when it is restricted to hardware people already own, rather than equipment they would have to buy.