Out-Of-Home Advertisement is a very big market which has stayed relatively untouched by technology. We want to create useful outdoor ads that display at the right time and at the right location to the right person !!
What it does
We created a context sensitive advertisement board based on the cognitive functions like visual recognition that understands the context as well as the users in a specific scenario. Therefore, offline ads especially the pricy ones at the airport, malls, etc. can be easily targeted just like its online competitors.
How we built it
With the cognitive functionalities of Microsoft Vision API at its core, we sense the surrounding and user in a particular scenario. We use Kinect to get the input key frames so that we can get the depth as well as extend our hack to interactions with our system. Apart from this, we created a web app wherein the advertisers can buy ads targeting specific users segmented by gender, age, etc. to better spend their marketing dollars. MongoDb is at the core of our product matching the ads with the users. Overall, the application is written in nodeJS and hosted in Azure!
Challenges we ran into
Identifying the key frames from a video stream to perform our image analysis was a big challenge.
Accomplishments that we're proud of
We created a working hack at the end of 24 hours !!
What we learned
We have to be careful about hardware hacks as they are not as reliable as the software ones.
What's next for Context Sensitive Advertisement
We really want to extend this hack to a product wherein advertisers can easily target outdoor advertisements as easy as buying ads from Google, Facebook, etc. with the same targeting capabilities as them !!