Fraunhofer develops a visual language model to achieve mutual understanding between driver and car

According to foreign media reports, optimizing the communication between vehicles and drivers according to the degree of automation is the goal of a research project carried out by Fraunhofer and other companies.

The researchers combine the sensors used to monitor the interior of the vehicle with the language model to form the so-called visual language model, which aims to improve the convenience and safety of cars in the future.

, photo source: Fraunhoff, “warning, if you continue to read now, you may feel sick on the winding road.

Please be prepared to drive alone for a while.

Safety comes first. “.To this end, the research teams of the Fraunhofer Institute of Optoelectronics, Systems Technology and Imaging Development IOSB (Fraunhofer Institutes for Optronics, System Technologies and Image Exploitation IOSB) and the Industrial Engineering Institute IAO (Industrial Engineering IAO) have partnered with ten partners, including Continental Group, Ford and Audi, as well as a range of medium-sized enterprises and universities to carry out the KARLI project.

KARLI is an acronym in German, which stands for “artificial intelligence (Artificial Intelligence for Adaptive, Responsive and Level-compliant Interaction) for adaptive, responsive and horizontally compatible interaction” in the future automobile.

“in the KARLI project, we are developing AI capabilities for L2 to L4 automation.

To this end, we record the driver’s behavior and design different human-computer interactions for each level, “explains Frederik Diederichs, project coordinator for IOSB at the Karlsruhe Fraunhoff Institute of Optoelectronics, Systems Technology and Image Development.

With different levels of interaction, depending on the level of automation, drivers need to focus either on the road or on something else.

They have ten seconds to regain control of the steering wheel, or in some cases they don’t need to intervene again.

These different user needs and the ability to switch between different levels of automation based on road conditions make defining and designing interactions suitable for each level a complex task.

In addition, interaction and design must ensure that drivers are always aware of the current level of automation so that they can perform their duties correctly.

, return to the first electric network home page >.

Link to this article: https://evcnd.com/fraunhofer-develops-a-visual-language-model-to-achieve-mutual-understanding-between-driver-and-car/

Like (0)
evchinaevchina
Previous July 3, 2024
Next July 3, 2024

Related Suggestion