Blog

Automatic AI Platform replacing manual & labor-intensive Data Annotation Processes

For several weeks we have discussed the several technologies that have helped us to stay ahead of the pack. In this issue we will address where Deep Learning fits into the whole concept of Artificial Intelligence.

To many, Deep Learning equals Artificial Intelligence. This myth has been perpetuated by the suppliers of GPUs as Deep Learning requires a lot of computing resources.

Deep Learning involves training a system with large data sets. This is expensive in terms of labour for labelling the datasets before the system can be trained. Also because of the large training datasets required the computing load is high requiring expensive GPUs. 

Deep Learning uses the memory function of the brain. It extracts common characteristics from large numbers of images of similar objects and looks for these in any new object that it sees. As such a Deep Learning system is only as good as the training data that it is fed. It will not accurately recognize an object if something very similar was not part of the training data. 

If one wanted to detect a dog, one could train the system with images of tens of thousands of dogs and when the next dog comes into the scene the system would recognize it. However what happens when you want to detect whether a person is about to commit suicide in a prison? How does one generate a data set with tens of thousands of images of persons about to commit suicide?

The human brain operates at a much more sophisticated level. Humans exercise many different types of thinking beyond using their memory function. Over decades of experience in AI, WildFaces has developed algorithms that emulate many of these different ways that human think. Deep Learning has its place in the AI schema. However, by converging it with the many other types of AI algorithms that WildFaces has developed, the system:

• Is very computing light and does not require GPUs.
• Requires a minute training data set greatly reducing the cost of labelling and training.
• Is quick to deploy as the user does not have to spend weeks and months training the system.

This is just one way that WildFaces systems do more with less – they can be less expensive to deploy than even simple recording systems.

Click HERE for video explanation of our unique AI Approach.  
You can also read more from HERE

Upcoming Event - Dubai Intersec 2023

Educational Seminars at our booth S1-A25

Joint Educational Seminar organized by WildFaces along with Intel and Lenovo will be run every 2 hours covering some of our following technologies and live demonstration:

• Smell Analytics 
• Anonymized Tracking across non-overlapping cameras
• Intelligent Traffic Management
• Intelligent Facilities Management
• Advanced Smart Cities

Please RSVP your preferred session HERE

About WildFaces:

⌂  www.wildfaces.ai    |    ✉  Contact Us

©︎ 2023 WildFaces. All Rights Reserved.