It all started with these pretty flowers standing on the dinner table. Everyone likes to receive flowers once in a while, right? I love flowers. They’re pretty to look at and they make me feel happy. The problem is that I’m not really a flower expert myself and had no clue what kind of flowers I got as a present. Too bad, because I would not know what to ask for the next time at the flower shop or what to order online. The first thing that came into my mind was: There should be an app for this!
After searching on two key words in the App Store without any result my attention was already somewhere else. Later on, I threw my ‘problem’ in the #ideas channel in Slack at work with the question if we could create an app that tells me the name of the flowers? Something like Vivino, an app that supplies wine-lovers instantly with the wine's rating, review, and the average price just by taking a photo of the winelabel. This app is an example of an awesome seamless user experience. I immediately got response from my Frog colleague Isabelle. She suggested to use the Plantifier app.
Great! I downloaded the Plantifier app, created an account and added the picture of the flowers. A couple of hours later I received an email. This nice person had identified my picture and declared that the flower was an Aster.
I Googled Aster images to be sure and confirmed that this was correct.
I like the fact that within this app the community answers. People help each other out and that’s a sweet thing to do. It is a social event and it gives satisfaction to the person who has the answer sharing his or her knowledge with others. They get the chance to teach others about their hobby or personal interest.
I must admit, I was a bit skeptical at first. I wondered if I would get the answer and, secondly, when this moment would be? I was depending on others that knew the correct answer and I also had to wait for someone to ‘feel like answering’ my question. This made me come to the realization that we lost our patience in this digitalized world. People want instant gratification. Everything needs to happen instantly and, in this case, it seemed like there was no time to wait for an answer. From this sense of urgency I shared in #ideas again: “.. I guess this could be automated scanning the pictures and checking on color, shape, similarities etc.”
Apparently my suggestion triggered the technical Frogs Eelco, Naos and Paul since they started to give their input on my idea. Paul explained that the cognitive services of Microsoft make us able to make applications more intelligent. It detects, groups and recognizes data according to visual similarities. And, here comes the great news, we actually have the image recognition in place and tested 98% accuracy when top 5 matches where shown. The system even specified my search result and gave back the flower name is a China Aster. How awesome!
This 'flower idea' real-life user experience shows a good example of machine learning opportunities. Of course, there are various types of “intelligent API’s” that offer image, video, speech, and text recognition that you’ve probably heard about before from other applications. But sometimes it is hard to connect a very specific user need to a technical solution that answers this need. I’m sharing this exploration because I think it makes a simple idea very tangible translating it to an app concept. In conclusion, the community is a powerful tool for knowledge sharing. But cognitive services can make us - make your app smarter more instantly. And it gives your mobile users the opportunity to discover and learn fast. And last but not least, now I know the name of the pretty flowers. Problem solved!
If you recognize sometimes thinking to yourself: 'There should be an app for this!' Please feel free to share these moments and reach out to us.