User gets to decide whether to drop more photos or view previous results
Provides Companies with more information on the Poster AI service
Photos are uploaded
User clicks capture reaction to have their picture taken by the web cam and then analyzed
The user selects whether they like or dislike the advertisement
The next picture comes up
The last picture uploaded goes through the same process as the first two
The results are calculated and displayed
Advertisements are a huge part of our lives, we see and hear them everyday, but how can one tell if their ad is effective? With Poster AI we solved this problem using ML to extract information about a user's emotional response to an ad while also gathering their input. This allows for a qualitative and quantitative analysis of the effectiveness of an ad. Whether the technology is used by a company to increase their revenue, or an organization to raise awareness for dangers facing youth, the applications of this tech is endless.
What it does
The web app determines how effective an organization's marketing is by analyzing a person's reactions to their advertisements. The platform is released online and users can earn credits for participating in the "surveys" from which companies gain data on their marketing success. It can also be used internally for company run user testing.
How we built it
We used Vue.js for the frontend to first gather data from the user. Then, using Azure's cognitive services we analyzed the emotions the user displayed while looking at the posters. This was integrated into a python and mongo DB backend. The results were then sent back to be processed and displayed in the results section.
Challenges we ran into
Building the application from the ground up was time consuming.
Accomplishments that we're proud of
We are proud of how we were able to implement our idea in the given time frame.
What we learned
We learned a lot using ML services and Microsofts vision api.
What's next for Poster AI
We see a future where this platform can used by big and small companies alike, so they can gauge audience engagement on their advertisements before releasing them, thereby saving millions.