Improved Multimedia Object Processing for the Internet of Vehicles.

Area of Science:

• Artificial Intelligence
• Computer Vision
• Edge Computing

Background:

• Intelligent edge devices combine edge computing and deep learning for conditional decision-making.
• Automated cars in the Internet of Vehicles (IoV) system act as data-source nodes.
• Real-time threat detection in smart cars relies on intelligent supervision cameras processing multimedia data.

Purpose of the Study:

• To achieve more accurate and rapid object detection in smart cars.
• To address real-time delays in data offloading and synchronization with cloud systems.
• To enhance the capabilities of intelligent cameras in automated vehicles.

Main Methods:

• Utilizing a cooperative machine learning technique for distributed computation.
• Slicing real-time object data among intelligent Internet of Things (IoT) nodes.
• Implementing parallel vision processing between connected edge clusters.

Main Results:

• Increased computational rate and improved accuracy in object identification.
• Achieved low latency and higher accuracy through real-time multimedia data objectification.
• Demonstrated responsible resource utilization and active-passive learning.

Conclusions:

• The proposed model enhances object detection performance in IoV systems.
• Cooperative machine learning and edge computing effectively reduce latency and improve accuracy.
• Intelligent edge devices offer a viable solution for real-time threat detection in automated vehicles.