The Affectiva Automative AI will analyse and determine the current state of the vehicle occupants through a series of cameras
Affectiva, a leader in human perception AI, has received six patents in relation to the detection of awareness and emotions of vehicle occupants.
The AI produced by Affectiva Automotive incorporates deep learning with computer vision via cameras placed within the vehicle itself. These cameras will analyse the current state of the driver and anyone in the vehicle in real time.
Co-Founder and CEO of Affectiva, Dr. Rana El Kaliouby, said:
“Over the last decade, Affectiva has continuously pursued new patents as we have pioneered and advanced the fields of Emotion AI and Human Perception AI.
The breadth and depth of our patent portfolio reflect our commitment to pushing the boundaries of computer vision, machine learning, deep learning and AI at the edge; and, is a testament to our leadership in defining the many creative and diverse applications of Human Perception AI that are shaping industries today and in the future.”
The main aim of this whole concept is obviously to do with safety. The majority of accidents happen due to human error with 95% of accidents related to this. This newly developed AI will look to reduce this percentage.
The AI will be able to instantly detect if a driver or its occupants are a little off in their emotions or states by detecting drowsiness or if they are distracted.
The following are the six patents:
1. Drowsiness mental state analysis using blink rate
An individual blink rate will be able to determine whether someone is experiencing any drowsiness, mental fatigue, lack of attention and/or experiencing cognitive load.
This information will be able to make the driver aware that they need a break, or even assisting the driver by activating braking and steering control.
2. Cognitive state evaluation for vehicle navigation
This involves a multi-modal sensing of a vehicle occupants' cognitive state. AI will analyse this through specific signals such as speech, yawns and snoring. Heart rate and respiration rate will also be analysed. Other factors are also analysed such as environmental conditions (temperature, weather, lighting)
3. Cognitive state-based vehicle manipulation using near-infrared image processing
This patent is focused on night time driving or when a car passes through a tunnel. The use of NIR cameras enables readable images under dark lighting. The cameras inside the vehicle will use deep neural networks that run on embedded systems.
4. Vehicle content recommendation via affect
This will provide personalised content to improve the transportation experience, enabling a new stream of monetised opportunities for vehicle manufacturers and advertisers.
5. Vehicle content recommendation using cognitive states
Similar to patent number 4. However, content is based on an understanding of a person's cognitive state determined from multiple sensors inside the vehicle. Content specific to the occupants will helpthem to focus, relax or calm down during driving.
6. Image analysis for emotional metric evaluation
The AI will collect individual's emotional responses to content that they will see on their mobile devices and in-vehicles. This will aid the individual to track their emotions over time (if they consent and opt-in) and serve as a fitness tracker for their overall wellbeing.
Affectiva has currently been awarded 39 patents and is still waiting for 25 more.
About AIBC Dubai:
AIBC UAE will take place on the 25th to 26th May, 2021, in the emirate of Dubai. The event will bring together key brands and individuals from the converging sectors of AI, blockchain, IoT, Quantum Tech to discuss and shape the future of emerging tech.
It is one of the leading events globally for blockchain, AI, crypto, and other emerging technologies, and gathers together an elite selection of delegates, policymakers and thought leaders from across the globe. Such international recognition has helped propel AIBC Summit become a favourite on the world circuit for emerging tech conferences and expos.