As the mental illness begin to be more accepted on society, we see a really big niche on chatbots as a solution to depression.
What it does
Our bot combines facial and linguistic features to treat depression by creating routines, improving your diet or suggesting spontaneous activities.
How we built it
The good box takes advantage of multiple technologies treat and monitor depression patients. We try to recommend different activities based on their emotional patterns, visual (computer vision) and linguistical (NLP). Particularly, we use opencv to check if there's someone on the field view to take a picture and send it to Google's Vision API, that runs all facial expression recognizance.
Challenges we ran into
We front the learning curve of install all the programs and dependences. Furthermore, we run into some conceptual problems like how chatbots work.
Accomplishments that we're proud of
Presenting a product with such a tangled systems is our proudest accomplishment. At first, and even on the finals of the hackathon, we lose the faith on the viability of the project, but we kept working.
What we learned
The most valuable knowledge is how to build a complete hardware/software solution from top to bottom.
What's next for The good box
We want to improve the empathy of the robot by using natural language processing and exploring better imaging techniques. Moreover, we would like to add a dashboard for the doctor to visualize the progress and automatize systematic tasks.