Cross-View Semantic Map Transfer for Autonomous Path Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving technologies face challenges in effectively transferring knowledge between sensors with different views, such as top-view and perspective-view sensors, which hinders accurate object recognition and path prediction.
Innovation Solution
A method involving map matching and knowledge transfer between rasterized semantic maps from top-view and perspective-view sensors, creating a motion flow map in a bird's eye view occupancy grid vector form, and using this information for path prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If knowledge transfer is attempted between top-view and perspective-view sensors, then path prediction accuracy is improved, but the complexity of the system increases due to the need for map matching and coordinate transformation between different sensor views
Solution Approach 1:
The patent introduces rasterized semantic maps as an intermediary representation that bridges top-view and perspective-view sensor data. The maps serve as a common coordinate system that enables knowledge transfer between sensors with different views, allowing motion flow information to be extracted and transferred without directly complex coordinate transformations between sensor frames
Solution Approach 2:
The patent segments the knowledge transfer process into distinct modules: map generation from sensor data, map matching between different views, motion flow extraction, and path prediction. This segmentation allows each module to be optimized independently and simplifies the overall system architecture by breaking down the complex task of cross-view knowledge transfer into manageable components
2Measurement precision
If multiple sensor views are integrated for motion flow analysis, then object recognition accuracy is improved, but the computational resources and processing time increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing sensor data into rasterized semantic maps that encode scene understanding information. This preliminary transformation organizes the data in a standardized format that facilitates efficient matching and motion flow extraction, reducing the computational burden during real-time processing
Solution Approach 2:
The patent creates simplified copies of the complex sensor data in the form of rasterized semantic maps. These maps are lower-resolution representations that capture essential semantic information while reducing computational complexity, enabling faster processing while maintaining sufficient accuracy for path prediction and object recognition
Data Source
AI summary
A knowledge transfer method allows for knowledge transfer based on images acquired from sensors with different views. The knowledge transfer method allows for knowledge learned from an infrastructure device including a top-view sensor to be transmitted to a vehicle including a perspective-view sensor, and the vehicle may predict a path for autonomous driving using the transferred knowledge. Rasterized semantic maps generated based on data acquired from each of the top-view sensor and the perspective-view sensor may be used to generate a motion flow result based on map-matching. The motion-flow result may be transferred to the vehicle for path prediction.