We've recently been having a lot of fun exploring the power of Microsoft Cognitive Services for a project we are working on. Microsoft Cognitive Services are part of Microsoft's Cloud offering (also known as Azure).
The services are split down into a number of APIs grouped into Vision, Speech, Language and Knowledge. There are some really powerful tools available which can all be accessed cheaply and easily as part of our Azure hosted solutions.
We're going to take a look at just a few of the services available.
The Face API provides a powerful face detection service. It can detect one or more human faces in a scene, and identify where the faces are, including the position of the eyes, nose and mouth. It is also able to estimate various properties of each face including age, gender, pose, smile and facial hair. You can try to Face API for yourself here: https://www.microsoft.com/cognitive-services/en-us/face-api
If the Face API isn't clever enough for you the Emotion API can go even further and estimate the emotional state of each person in the scene. The emotions detected are anger, contempt, disgust, fear, hapiness, neutral, sadness and surprise. These are the emotions which are believed to be cross-cultural and universally commincated by facial expression. Again you can try this API for yourself with your own images here: https://www.microsoft.com/cognitive-services/en-us/emotion-api
The Content Moderator tools allow automatic moderation of image, text and video content. This includes automatically identifying and flagging profanity, malware, phishing URLs and personally identifiable information. If you have a service or system that processes a large amount of user supplied data then running it through the Content Moderator API can quickly help filter out content that shouldn't be passed on for other users to see and share.
The Cognitive Services include a range of speech recognition tools, both general purpose speech recognition and trainable/tunable bots that can be tuned to specific user's voice patterns and acoustic environment. These speech recognition tools can work great on mobile devices saving users from having to use a small screen UI.
The Linguistics API takes input text and parses it to try to extract meaning and concepts automatically. Many sentences can be broken down into constituent parts automatically and the basic meaning of the sentence can be extracted and processed without any human intervention.
One of the most powerful and potentically profitable APIs is the Recommendations API. This service can learn from your historic transactions and then make recommendations based on previous shopper behaviour. What makes this service really powerful is that it isn't just a "People who bought A also bought B" style of recommendation. It can also take into account your shopper's entire shopping history, page view activity, and use external knowledge sources to create connections and recommendations.
These are just a few of the ever growing suite of cognitive services which are available as part of Microsoft's Cloud offering. We're having a lot of fun exploring these and finding ways to make use of them in our software. If you think your systems (or new project) could benefit from the power of some of these services please give us a call to discuss.
