Monday, November 26, 2018

Azure Cognitive Services in Containers

Do we really need data scientists to create a solution using machine learning? 
Well, previously, yes. It’s a must. 
But now thanks to Azure cognitive services, those can be done by a developer. 
If a developer wants to take the advantage of AI and features cognitive service provides which is vision, speech, search, text processing, language understanding this will be the best way.
Anyone who can write a simple program, means who writes basic code can use this feature and create AI models using AML. But the problem is scaling. Most of the time model developers facing this problem and now there is a solution for that too. Container Support for Cognitive Service. It is much easier to use and just build and deploy.
This allows developers to build big AI systems that run at scale, reliably, and consistently in a way that supports better data governance.
For better data governance, developers need to keep AI systems scale, reliably and this allows that to developers easily.
Ex: A School System
First things first. There may be some entities for main actors for teachers and children. But later on there will be many other entities will be there such as, attendants, notes, news etc. So while its grow its hard to maintain and scale it once it’s on production.
For this, can use these kind of solutions (Cognitive Service) to keep track and scale easily.
When it comes to large scale apps, it’s so much easy to maintain the app as well.

There are 4 key capabilities within Azure Cognitive Services
1. Text Analytics Container
  • Key Phrase Extraction          

It creates a key word by searching words and return those key word by using AI.
Ex: “what are the nearby coffee shops to go?” (Coffee shop, nearby)
  • Languages 
Supports up to 120 languages
  • Sentiment Analysis
This work as raw text analysis and returns a value in between 0 and 1 where 1 is the most positive

2. Face Container

Face detection is a most valuable fact now days which each and every device has the feature. So the API detect the face, verify and most of the time it is trained to detect emotions even. It is a common framework runs underlying and user can easily set up security, storage and logging for his container.

3. Recognize Text Container

OCR (Optical character recognition) has improved in here and this can detect text from 120 languages and of various objects and backgrounds. 

4. Custom Vision Service support for logo detection

Custom Vision Service will add support for logo detection, allowing business to create their own logo detector quickly and easily. Logo detection is a specialized type of object detection suited specifically for logos that can be small, skewed, or obfuscated within a larger picture, for example on the sidelines of a soccer match, on a building sign in a cityscape, or on a scanned form. Now you can build your own logo detectors to help search and locate their logos in their media libraries or to generate analytics for their social media feeds. 
While supporting to each and every thing, why not object logos? Users can create their own logo detectors which to identify their businesses. It is a bit advanced algorithm and it will detect specially logos and brands. Those logos might be blurred, small, fuzzed, no matter what it will detect it. 
Read full article on Azure Blog: 

No comments:

Post a Comment