Combining Human with Artifical Intelligence
Computer vision is much faster than humans in some recognition tasks. However, the accuracy depends on the availability of training datasets, the quality of image annotations and the adaptation of algorithms for specific tasks.
The human brain is a powerful information processing machine. It outperforms any computerized technology when it comes to dealing with novel, abstract, or fuzzy definition of tasks, due to incredible learning efficiency and cognitive flexibility.
By connecting humans and machines, InnerEye combines the best of both worlds.
He can scan 10,000 images per hour
This breakthrough technology made by InnerEye combines the skills and expertise of the human user, cutting-edge neuro-technology and state-of-the-art machine-learning and signal processing algorithms.
The novel active learning framework for training deep neural networks for visual recognition is using the information measured directly from the brain of human experts.
By presenting selected images at a fast rate up to 10 Hz and classifying brain responses to generate labels, it becomes possible to improve the accuracy of the model, as well as to reduce training time and data.
InnerEye's technology by-passes the need to record overt responses from the user, reading the visual recognition signals directly from the user's brain, and combining it, together with the visual data into a unified Artificial Intelligence system.
Deploying for Aviation Security
Today maintaining passenger security is a non-efficient and low-quality operation. 1 x-ray machine needs 1 security officer and the outcome is only a low threat detection accuracy.
With InnerEye's technology there's a so called A-Team approach with 1 security officer covers 5 x-ray machines while detection accuracy is much better.