Chad Hart describes his winning TADHack leveraging web-based computer vision API's and WebRTC and how this process can help facilitate product development without requiring more professional developers.
5. How it works
Eye App Monitor App Alerts on Any device
TURN network
WSC
OCSG
WebRTC Control
SMS
Computer Vision
SMS
Calls
Email
6.
7. What the apps look like
Alarm go off
after a trigger
Types of triggers
“faces”:[{“age”:
4}..]
Project Oxford API
detects a 4-year old
face
“Remote Eye” –
a remote camera
is in my fidge.
The Monitor
receives this
video and
processes the
stream
9. I’m not a professional developer. I’ve
never been one and I don’t want to
be one.
WebRTC & modern comms APIs mean
“product” people can make & iterate
on their own prototypes before
handing-off to the professionals.
10. There’s not enough hack-a-thons!
I only used a subset of one object type
There dozens of others and more being
added
This doesn’t include the Project Oxford
face detection or speech API
Everyone one of these squares could be
an app
11. There’s not enough hack-a-thons!
There are lots of other exciting computer
vision, natural language processing, an
emotion analysis API’s
These are all great candidates for RTC
mash-ups
12. the web-way is the only way
WebRTC let’s you mash up RTC with other web API’s. While other RTC API’s exist,
none provide the ease of development and convenience of using a browser that
WebRTC does. I would not have been able to pull off this hack without WebRTC.
13. For more
• TADHack blog: http://blog.tadhack.com/2015/07/20/marketing-guy-
won-tadhack-thirdeye/
• Demo video: https://youtu.be/RwFlVJco2ME
14. About Me
Chad Hart
Product Consultant
& Chief Editor, webrtcHacks
https://webrtcHacks.com
@webrtcHacks
@chadwallacehart
chad@chadwallacehart.com