Inspiration:
I want a soundtrack to my life. When I am listening to good music, everything in life is better. Unfortunately music isn't playing all of the time because 10,000 other things require my attention.
Music and lights have the ability to completely change a persons mood.
A good DJ can change the mood of an entire venue, they gauge the mood of the crowd, use their advanced music knowledge then rinse and repeat.
Productivity studies have shown lighting can increase productivity and invoke different moods / thinking patterns https://www.entrepreneur.com/article/243749
Over 70% of communication is non-verbal. Luckily computers are good at understanding non verbal communication as this is more algorithmic / mathematical than words and voice.
The future of IOT and technology is that of automation and emotionally aware machines.
Instead of us controlling our environment to change our mood, machines need to start doing it for us. Our time and mental capabilities are limited.
What it does:
Changes lights and music based upon your mood without you having to do anything.
You can choose to set a desired mood and it will learn what music is/isn't effective at bringing you to your desired mood.
Stretch goal:
- The user can design the moods of their day, i.e. I want to be productive from 10-12, relaxed from 12-1 etc.
Mood is quantified using mood recognition software FaceReader from Nodulus.
Technologies being used:
Raspberry Pi
I have a raspberry pi 3 we can use.
Arduino:
Alorium Tech XLR8: This runs the Arduino IDE. This allows us to Change pixel brightness without corrupting color data. One of the biggest weaknesses we ran into last year with the Juke Box Backpack was updating brightness real time via sound level would sometimes cause the entire program to crash.
Facial Detection / Mood detection:
FaceReader by Nodulus I have spoken with the company and received a provisional developer license that they will extend as long as we need.
I am also working on getting a different software from Emovu
Spotify
Neopixels
I sell on amazon and buy from china. As a result I have over 500 meters so we are set.
Speakers
We need speakers. I don't posses speaker knowledge so some help will be needed to purchase an effective speaker driver.
Talent Needed / Desired:
Data skills
We will be receiving and gathering mood data every second, we need deeper insight into the second to second mood to derive the overall mood of a user.
Programming Raspberry Pi
Linux
Web API
Microcontrollers
Android or IOS
3D Modeling / CAD
If we make it into a portable bluetooth speaker we will need to build out a design.
Speaker Design
It needs to be functional, cool looking and sounding.
Electrical
Electrical engineers will be much loved!
Lighting display
If someone has experience taking images and pushing them to LED displays or creating sequences / patterns it would be helpful. I have quite a bit of experience but I have never done graphical / crazy visual displays.
Challenges expected:
Determining when to change music based on the persons facial expressions / mood.
Image quality / lighting conditions
Storage space and operating capabilities with raspberry pi
What we'll learn:
How accurately we can predict mood with this software
How often to change music based on persons facial state.
What's next:
Finding an applicable use case.
Interviewing potential customers to further refine design
Built With
- arduino
- facereader
- neopixels
- raspberry-pi
- spotify
Log in or sign up for Devpost to join the conversation.