Hand Gesture Recognizer

Inspiration

Driven by the belief in inclusive technology, our inspiration stems from a desire to empower individuals with disabilities, providing them with a newfound means of expression and connection. JavaScript Hand Gesture Recognition is a testament to the transformative potential of technology, breaking barriers and opening doors to a more accessible and inclusive digital world.

What it does

The project utilizes TensorFlow in real-time to recognize hand gestures on web interfaces created using JavaScript, CSS, and HTML. This functionality provides users with an additional means of communication, particularly benefiting those with disabilities. It translates recognized hand signs into words and emojis, enhancing accessibility and enabling a more intuitive and inclusive user experience on the web.

How we built it

We built this website using JavaScript, and utilizing tensorflow to retrieve a 2D array that contained the presition of the fingers of a hand. Using these positions we calculated the distance from the palms and determined whether or not certain fingers were extended. Using these constants we distinguished the different hand gestures being made.

Challenges we ran into

We had a difficult time differentiating the gestures as many of them had the same style. For example, differentiating a horn sign versus love in sign language resulted in a mixup and only returning the horn sign. We also had a difficult time adjusting the program to compensate for distance from the camera and handling certain gestures.

Accomplishments that we're proud of

We accomplished a program that sucessfully recognizes certain hand gesetures and we're happy with the temporary UI that we made to demonstrate the program. We are also proud of the collaboration between team members and the goal we accomplished in the time crunch we faced.

What we learned

We learned the difficulties of creating programs within a group and having to readjust the code to fit the changes each of us made. We also learned the capabilites of javascript and how utilizing API could aid in accomplishing our goals.

What's next for Hand Gesture Recognizer

In the future, if we were to continue working on this, we'd definitely try to incorporate other hand gestures and attempt to create a more accurate hand gesture recognition system. Perhaps we'd update the current model to account for angles and rotation. We'd also fix the UI and add more features to the program.

Built With

• css
• html
• javascript
• tensorflow