Soon Will Remember

• 

  project diagram

• 

  one object result

• 

  two objects result

Inspiration

There are a lot of people with disabilities who is seeking to help. A handful of gadjets were invented for most common impairments, e.g. blind, deaf, injured people. Our idea specializes in a temporary disability called Aphasia, it happens after a heart stroke or head injuries. How does it look like? Person want to say "I want to eat" and instead of that he says "I banana fish meat" or sometimes "I want to ehmm ... you know.. putting food in mouth ... ". In former one, he thinks, that he said just what he wanted and do not understand why everybody is staring at him. In latter one, he can not remember basic words. We want to help them go through it easier.

What it does

In our MVP image from camera is being processed, and trying to understand where person is and what does he want to say. So he can read phrases or just show them

How we built it

We have used Yolo v11 nano to process image, it a very tiny model which can be ran on micro processors. We used Gemini 2.5 as a brain to process the results and give response.

What's next for Soon Will Remember

Our idea could be implemented in modified GoPro as Yolo weighs only 5mb and brain is located in a cloud which only requires internet connection using Saily's Esim for example

Built With

• gemini
• python
• yolov11