Inspiration
In everyday its a common thing signing without reading. While this is a bad practice we sign it anyway. We have no way of knowing if the signee read it or not. The situation is much worse when the document is shipped via post, signed and post back. While researching on the edge camera face landmark recognition. I see a new way of improving signing process online.
What it does
With help of camera it tracks your eye movements. We expect your eyes to move left to right or right to left while reading documents and blink while looking at the document. Since it runs on Edge device you have no problem with privacy laws like GDPR.
How we built it
Its designed like a captive portal. When user logins he is met with a document to read and a button to click
Challenges we ran into
While the technologies used are there, integrating them and having a working workflow is hard.
Accomplishments that we're proud of
Get my feet wet for edge AI on the web using tensorflow.
What we learned
Deployment of tensorflow AI models on web on edge.
What's next for You did read it. Right ?
I will use this face landmark detection with RAG with LLM. Also I will use it in a project so that people will paralysis can communicate.
Built With
- c#
- javascript
- pdfjs
- tensorflowjs
Log in or sign up for Devpost to join the conversation.