Gesture typewriter

A typewriter that recognizes gestures and turns them into words. What do you need that for, you ask? For a large group of people in our society, this is indispensable. Namely for: deaf people who cannot, or have difficulty, reading and writing. Because how else do you get in touch with hearing people? Or how do you send an email or message to your loved one? And make no mistake, this group of people consists of an estimated 4,000 people. In addition, this tool is also useful for all other deaf people. The tool makes communicating with hearing people a lot easier.

Can you help us?

We are very excited about this project and are working hard for a quick result. You can help us to speed up and improve this process. If you read this message and think you can contribute. Do not hesitate to contact us. We can use all the help, but especially from people with the following knowledge:

Computer vision and/or machine learning. A person with knowledge of these subjects may be able to speed up and optimize the process. We have knowledge of these subjects ourselves, but we do not have much experience with similar projects. Do you have that? Then contact us.

Experts by experience. Are you deaf or hard of hearing? Or do you work a lot with this group? Then you can probably help us. We would like to talk to the group for whom we are creating the tool. This allows us to better tailor the system to their needs. If you think you can help us, please contact us.

How did this come about?

One of our employees was talking to a personal assistant of deaf people. During this conversation, the assistant told us that many deaf people have a delay in their language development. To date, NGB (Dutch Sign Language) has not been recognized as a language in the Netherlands. In fact, sign language has been suppressed for years. In the past, children were not taught sign language and were not allowed to use this language among themselves. Because deaf children could not communicate well, they suffered a significant language delay.

Fortunately, deaf children now receive appropriate education, but there is still a large group of deaf people who cannot read and/or write. For these people we are now working on this project. Of course, we do this completely without obligation and without subsidy.

From problem to idea

Due to the language development deficit of deaf people, the bridge between that group and people with good hearing is large. We want to reduce this bridge. The majority of the Dutch population does not know sign language. That is why we want to develop a tool that converts signs into a text document. In this way, these people can write letters or emails themselves. Or even have a live conversation with someone who has no knowledge of Dutch sign language.

Where are we now?

At the moment we are still at the base of the project. We are experimenting with computer vision. Which points do we need to measure to recognize gestures. We have already been able to automatically recognize cars, motorcycles, cyclists and pedestrians. The facial recognition also works well. Now we are working on recognizing fingerspells (the alphabet in sign language). Once we have fingerspells figured out, the step to other gestures will be less big.

Why are we doing this?

Initially, we do this mainly because we think it is important to actively make a positive contribution to society. We heard about this problem and found it a great challenge.

In addition to the social part, we do this to expand our own knowledge. Computer vision and Machine Learning are becoming an increasingly large part of our current society. That is why we want to know all the ins and outs of this. If we work with these principles in practice ourselves, we can advise our customers better later.

If you can help us with this project, please contact us.