Gesturely needs your camera to see your signs, and your microphone to hear what the other person says back. Your video is processed in the moment and never stored.
A Deaf person signs. Gesturely turns it into English on the screen. The hearing person replies by speaking or typing, and it appears as text. No interpreter to book. No notepad passed back and forth. Just a conversation.
A camera watches for signing and clips each sign as it happens — no button to hold, no start and stop.
Each sign is recognized on its own, then a language model stitches the glosses into fluent English sentences.
The hearing person speaks or types. Their words appear as text on the same screen, so the exchange stays in one place.
Not sure? Pick the first one — nothing is saved and you can change it later in Settings.
Sign language is a full language with its own grammar — but almost nobody on the other side of the counter speaks it.
Nothing to install and nothing to set up. Sign, and your words reach the other person in English — on the screen, and aloud for them. What they say back appears here as text.
One phone between two people, taking turns. It runs on whichever phone is present — the Deaf person does not have to be the one carrying it. Nothing either of you says is saved.
Face to Face works by passing one screen back and forth: you hold it, then turn it to face the other person. That only works on something you can hold in one hand.
Open gesturely.ai on your phone and come back to Face to Face.
Only used for signing — it sets which words Gesturely expects.
Choose where you are first. In this mode they’ll be signing to your camera, so Gesturely needs to know which signs to expect.
Walk through a pre-written conversation between a deaf signer and a hearing person. Follow along step-by-step to see how Gesturely translates signs into spoken English in real time.
Upload a pre-recorded signing video — a story, a monologue, a recorded conversation turn — and get back a multi-sentence narrative transcript. Each sign is recognized individually, then an LLM stitches them into well-formed English sentences.
Work through a lesson's vocabulary in three phases — single signs, short phrases, then whole conversation turns. Each phase unlocks when you pass the one before it, and your progress is kept per lesson.
Your progress is stored against a learner code — no email, no password. Without one, this sitting counts as you go but is lost when you close the tab. Once you have a code this device remembers it, so your progress comes back on its own next time. You only need to type it in on a different device — so write it down.
Students can practise alone all they like and still have no idea whether a sign came out right. Gesturely watches them sign and tells them what the model recognised — the one kind of feedback that currently needs you in the room.
It is free while it is a preview, and what we want back is which of the missing things you would actually need. That list is what gets built next.
A kiosk is a tablet that stays at your counter, locked to one screen. Setting it up somewhere else would save settings that belong to a different screen.
Open Gesturely on the tablet you plan to use — upright or sideways, whichever way you will mount it — and come back to Setup Kiosk.
Gesturely becomes a single screen your visitors can sign to. You make the decisions once, now — then nobody has to touch a setting again.
If you have a kiosk code from another tablet, enter it and this one picks up the same setup.
You don't need any of this to try Gesturely. Setup is only for leaving a tablet out for other people to use.
Every sign the models know, with a short reference clip for each. Watch the reference, then practise against the model to check yourself. 0 signs available.
Help improve Gesturely by recording your own clips of a sign. Pick a sign that has a reference video, mimic it, then review and save — your clips become training data.
🔒 Only the pose skeleton is saved — your recording is used to extract the pose and is then discarded. Your actual video is never stored.
Configure how Gesturely captures, recognizes, and presents the conversation. Changes are saved on this device.
I’m Deaf.
This phone turns what you say into text I can read, and shows you what I want to say.
Please talk normally when the screen says TALK NOW.
Hold this up for them to read, then tap Start.
Choose where you are first, so Gesturely knows which signs to listen for.
Whichever you pick comes up big, for them to read.
Hold the screen up, then read their reply.
I want to understand you.
This phone reads sign language. I’ll point the camera at you — please sign one thing at a time.
Then I’ll turn the screen round, so you can read my reply and check what it read.
If you’d rather not, just shake your head.
Hold this up for them to read, then tap OK.
Hold the screen up for them to read.
Let's break down how Gesturely translated the signer's signs into English sentences.
Gesturely is powered by SignBridge, our peer-reviewed ASL-translation research.