Inspiration
To aid those who, for whatever reason, were not endowed with the same critical sense of sight as the rest of us,. Many people who are visually impaired confront daily challenges since they know they will never be able to function independently in society. Despite their immense and admirable courage and resolve, these folks must confront the grim reality of the ever-prolonging night. Sensifi was designed with the hope of assisting people with strong personalities and we are proud of our stand.
What it does
Sensifi builds a 3D dome for the visually impaired. We accurately forecast and segment all obstacles in the vicinity of a walking individual. As soon as the software detects any of these hazards in the immediate area, it will suggest a course of action that will take the user safely through them all. When using the map function, the user may rest assured that they will continue to be guided along the most direct route possible, minimising their chances of encountering any potentially hazardous situations along the way.
How we built it
It's a smartphone app, and it was developed with the help of the React Native framework. TensorFlow is used to construct the DL models, and Flask is used to access the DL model's APIs. There are a number of other external libraries used as well.
Challenges we ran into
The most difficult task that we had to complete was determining which features of the surrounding environment were significant enough to be considered potentially dangerous and so needed to be avoided.
Accomplishments that we're proud of
We are extremely proud that today we have been able to deliver software that can be used to benefit society. This project will help hundreds overall and could possibly make someone's life easier. We are also pleased with the accuracy we have attained from our models.
What we learned
We as a team learned the true importance of time management. We learned how to integrate DL models to our native apps and it's was very insightful.
What's next for Sensifi
Currently our dataset that was used could be a bit better if this project were to be rolled out as a product.
Built With
- flask
- javascript
- keras
- python
- react-native
- tensorflow
- yolo
Log in or sign up for Devpost to join the conversation.