We are already connected to AI in one way or the other with the advancement in technology- be it Siri, Netflix or Uber.
Yes, the technology is in its initial phase and more and more companies are investing resources in machine learning, indicating a robust growth in AI products and apps shortly.
The following statistics will give you an idea of growth!
- In 2014, more than $300 million was invested in AI startups, showing an increase of 300%, compared to the previous year (Bloomberg)
- Between 2018 and 2019, organizations that have deployed artificial intelligence (AI) grew from 4% to 14%
- More than 40% of enterprises will create state-of-the-art digital workers by combining AI with Robotic Process Automation by end of 2020 (Forbes)
- By the end of 2020, “customer digital assistants” will recognize customers by face and voice across channels and partners (Gartner)
- Artificial intelligence will replace 10% of American jobs by the end of the decade (Forrester)
- 15% of Apple phone owners’ users use Siri’s voice recognition capabilities. (BGR)
While basic AI has existed for the past 20 years, it is only in the last few that the development of these technologies has spurred on exponential growth and interest in the industry.
In 2012, AI research made a big breakthrough in the shape of neural networks.
Neural networks are way more powerful than the previous learning methods and provide the best results in numerous real-life problems. Their development is backed up by advances in computing power, which allowed running a huge number of parameters and the complex, data-heavy learning process.
Every time there is a new app, it is easy to imagine how recommendation engines, sentiment analysis, image recognition, and chatbots could improve the functionality of the app. All of these applications have been made possible or greatly improved due to the power of neural networks.
How are AI products built?
When it comes to building products and services based on AI, companies commonly go with one of two approaches. They either use plug&play APIs and machine learning models or develop their own, specialized solutions.
The most preferred choices for many organizations are off-the-shelf APIs. The leading providers make them easily available attracting more developers to use their platforms and very affordable. Amazon Rekognition, Google Vision, or Microsoft Cognitive Services APIs are becoming advanced and powerful due to this.
These tools are not flawless. Many APIs enable text recognition on images but are not accustomed to specific types and formats. For instance, they cannot recognize names and surname on someone's Identity card. At ONE BCG, we understood the gaps and strived through to develop a robust face recognition system using deep learning.
The other approach AI-based companies take is building custom solutions. The main reason for the growing popularity is the number of machine learning models available for training different models. But these models have certain drawbacks when it comes to practical implications.
ONE BCG has strived through to develop a robust Face Recognition system that would aid organizations to authenticate folks in their premises. In fact, at the moment we have procured 90% accuracy using Deep Learning which was initially based on template matching (K-Nearest Neighbours Algorithm). KNN is one of the basic algorithms which comes under machine learning which is one of the powerful yet simple where it learns the pattern in a continuous form. But it has some drawback as its a slow algorithm takes time to detect faces and in a real-world scenario, it's not applicable
Definitely, it's not possible to make this wonderful technology without logic. There are various algorithms and approaches to implement this technology to obtain desirable results.
In the next blog, we will discuss at length the technique that worked wonders for us.
No comments: