Overview / Usage

Well, the modern world has given us much good stuff. But, the most important thing we all look for is good sleep! The sleep rate of the people has been reduced by 20%, reports say. A shocking statistic about insomnia says that 30% of the people have this. To get sleep or induce sleep, we all almost rely upon Music. Yes, it gives us good sleep.

There comes the problem, while music is helping us with good sleep being induced, there is a negative side to it. Most of us, 99% of us, keep the headphones on and music played even after we sleep off. There is no technique to switch off the music and to let you sleep without any disturbance. Listen, We get a disease called Noise-Induced Hearing Loss if too much input goes to the brain and ears when not required.

Music becoming / turned noise is our concern. We through our approach do the following:

1. Suggest nice songs which can induce sleep (This is based on the Ragas which can provide you sound sleep, Ancestors have provided this data for us!) List of Ragas (Carnatic /Hindustani ) Music which helps to induce Sleep :

Raagas:

Nilambari Bageshree Darbari Kanada Aboghi Raga Bihag, Karaharapriya Harikamboji Hindolam Bahar

1. We detect the moment you sleep and we slowly cut off the music and stop playing music thereby not disturbing your sleep and not damaging your ears.
2. The music recommender system is based on ML techniques and we get your stats and recommendations as well.

Most importantly, Music is divine medicine. It should not let you take medicine! We are here to get a perfect solution!

Methodology / Approach

We have used API - Activity Recognition Client - This is a machine learning-based API which shall track the activities and give a confidence score for the activities. This is very efficient and it consumes very little power. Certainly, it is very very useful in terms of power consumption as this API uses PENDING INTENT which will activate the sensors based on the change in activity and hence they need not be active all the time.

The activities which are tracked by the Activity recognition client are:

1. Still (Reference number is 3, we shall show you this in the demo)
2. On Foot
3. Walking
4. Running
5. In-vehicle
6. On Bicycle
7. Tilting
8. Unknown

We take the screen activity and Activity recognition client confidence score together as parameters to gradually reduce the volume of the music and to switch off the music without disturbing the user.

The volume shall be reduced from 100 to 75 to 50 to 30 and to zero.

The APP workflow:

1. We first have an interface where the user enters the application using his name and age. There is no authentication for it as anyone can use our app.
2. We have a "Remember Me" feature so as to remove the burden of typing name and age every time for a day to day user. In the same layout as 1, we have a button to enter into the admin interface.
3. Here we have authentication with username and password because not everyone is an admin.
4. Inside user activity, we have the media player and the list of songs with the song name, artist, raaga, and duration. The user can select a specific song (if he feels to) or can select any song generally. We have pause, play and skip to previous, skip to next options. We also have the seek bar to seek to a specific part of the music.
5. Our app plays music for at least 10 minutes and then detects the user if he has slept or not and if slept, we'll gradually reduce the volume and switch off the music and close our app.
6. Inside admin activity, we have to select the song from file explorer, mention the song name, artist and raaga and we have to click on upload. The songs will get uploaded in Firebase database and storage. We have a progress bar to show the upload progress.

Technologies Used

Android

​

​