World Model Scene Generation for Realistic Autonomous Driving Tests

Solution Overview

Problem

Conventional intelligent driving function tests rely heavily on real-world road testing, which is resource-intensive and requires skilled drivers, and existing scene simulation methods fail to generate scenes with sufficient realism.

Innovation Solution

A method and apparatus for information processing that acquires initial scene information, generates target scene information using a world model, and includes a device with a processor and memory to execute instructions for simulating realistic driving scenarios by enhancing scene realism through 3D modeling and data processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If real-world road testing is used for intelligent driving function tests, then test reliability is improved, but resource consumption increases and skill requirements increase

Engineering Contradiction:

Improvetest reliabilityVSAvoidresource consumption

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent creates virtual copies of real-world driving scenes through scene simulation technology. The simulation system reconstructs road environments, vehicles, pedestrians, and traffic conditions into digital representations that can be repeatedly tested without consuming physical resources. This copying approach maintains test reliability while eliminating the need for actual road testing resources.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical physical testing system with a computational simulation system. Instead of using actual vehicles and roads for testing, the system uses computer-generated environments and virtual sensors to simulate driving scenarios. This substitution eliminates resource consumption while maintaining testing capability through software-based simulation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If real-world road testing is used for intelligent driving function tests, then test reliability is improved, but driver skill requirements increase

Engineering Contradiction:

Improvetest reliabilityVSAvoiddriver skill requirements

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent creates virtual copies of real-world driving scenes through scene simulation technology. The simulation system reconstructs road environments, vehicles, pedestrians, and traffic conditions into digital representations that can be repeatedly tested without consuming physical resources. This copying approach maintains test reliability while eliminating the need for actual road testing resources.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical physical testing system with a computational simulation system. Instead of using actual vehicles and roads for testing, the system uses computer-generated environments and virtual sensors to simulate driving scenarios. This substitution eliminates resource consumption while maintaining testing capability through software-based simulation.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Quantity of substance

If existing scene simulation methods are used, then resource consumption is reduced, but scene realism is insufficient

Engineering Contradiction:

Improveresource consumptionVSAvoidscene realism

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent implements feedback mechanisms in the scene simulation system by continuously comparing generated virtual scenes with real-world observations. The system uses sensor data from virtual sensors to validate and adjust scene parameters, ensuring high realism. Feedback loops allow the simulation to self-correct and improve the accuracy of generated environments.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent employs parameter changes to enhance scene realism by dynamically adjusting multiple parameters including lighting conditions, material properties, geometric dimensions, and environmental factors. The system modifies these parameters based on input data and simulation requirements to generate highly realistic scenes that accurately represent real-world conditions.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20260011071A1Method for information processing, apparatus, device, computer-readable storage medium, and computer program product

Publication Date: 2026.01.08 BEIJING CO WHEELS TECH CO LTD

AI summary

The present application provides a method for information processing, an apparatus, a device, a computer-readable storage medium, and a computer program product. The method includes that: initial scene information of a to-be-tested road section is acquired, where the initial scene information includes a scene image or a point cloud of a scene; and target scene information of a target vehicle traveling on the to-be-tested road section is generated according to the initial scene information by using a world model.