Often, trips to the psychologist or the counselor turn into discussions about what we feel at that particular moment. What's even worse is that elderly and those near the poverty level either aren't getting their mental health checked up enough, if at all. This is evident because mental disorders are most common in both of these groups: the elderly and the socioeconomically disadvantaged. Even when they do visit a mental health professional, the presentation of themselves to the mental health professional may only represent how they're feeling in that day. This problem of horridly low frequency of mental health tracking amongst these groups is what we sought to fix with our app FeelShot. FeelShot is a free mobile application that lets anyone track their daily mood and feelings so they can have peace of mind and also have a record of such informtion to show their mental health professional.
What it does
FeelShot allows the user to take a picture of their face (selfie) in which an image analysis outputs how they are feeling today. 'Face Rectangles' are detected by the Microsoft Cognitive Services Face API and this result is used to run a mood/emotion analysis that outputs a whole number percentage for 8 key expressions: anger, contempt, disgust, fear, happiness, neutral feelings, sadness, and even surprise.
FeelShot also allows a user to record voice input about their daily routine, from which is also analyzed the deciphered text is analyzed. Accessibility is a huge issue, and for the aging population physical problems like tremors alone make it hard to type on a keyboard, which is why we used voice to text. This voice recording is converted to text using the Google Speech Recognition API. The resulting text output is then subjected to sentiment analysis using an open source text sentiment analysis (http://sentiment.vivekn.com).
The resultant dataset of emotion analysis (from picture and voice-to-text) spanning over a certain period of time is visualized using an Android-plot library and different kinds of charts and graphs are made available to the counselor or mental health professional for better understanding of the subject's psychological behavior.
Firebase is used for authentication of users using Google services (accessibility), and is also used as a database for every user
How we built it
Android Studio is used as an IDE to develop the Android App which performs the above mentioned tasks using various libraries and APIs from Microsoft.
Challenges we ran into
- Synchronizing all the APIs to get the desired output was extremely difficult.
- Using Firebase was a very good learning experience
- Maintaining the data in Real-Time.
- Data Visualization
Accomplishments that we're proud of
- Making a dynamic Real-Time application that could be helpful for psycho-analysis of the populace.
What we learned
- Team Collaboration
- Learnt FireBase
What's next for FeelShot
- Creating a Web Console for the Counsellor to manage and visualize the patient data.
- Be able to suggest recommended prescription using Deep Learning technologies like IBM Watson.
- Expanding it to all the platforms like iOS, Windows and Web App.