System, Method, Computer Program, and Computer-Readable Medium for Generating Annotated Data

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention describes a system that generates realistic, fully-annotated data for the detection, classification, and segmentation of objects and their properties using simulations and artificial intelligence (AI). The system digitally simulates sensors (such as radar, lidar, camera, or ultrasound) in a virtual 3D environment with both static and moving objects, then automatically provides detailed labels and attributes to the data. These high-quality, annotated simulation datasets can be used to optimally train AI models for object detection and signal improvement, and can replicate or surpass the capabilities of real sensors.

Use CasesContent extracted from patent full text and abstract with AI.

• Training AI for autonomous vehicle perception (object detection, classification, and segmentation).
• Testing and improving the performance of automotive radar, lidar, camera, and ultrasound sensors in virtual environments.
• Developing and benchmarking AI-based signal enhancement techniques (e.g., super-resolution, noise reduction).
• Medical applications such as monitoring vital signs, movement analysis, and patient tracking using simulated sensor data.
• Training smart home systems for human presence and movement detection, fall monitoring, or energy management.
• Industrial robot navigation and object detection in logistics or manufacturing settings using synthetic sensor data.
• Security and surveillance applications, such as traffic or parking lot monitoring with realistic sensor simulations.
• Research and development of new sensor technologies and sensor fusion systems before hardware deployment.
• Military and defense applications including autonomous drones and robotics with advanced sensing and recognition capabilities.

BenefitsContent extracted from patent full text and abstract with AI.

• Automated generation of large quantities of realistic, perfectly annotated training data, greatly reducing manual labeling effort.
• Enables the training of AI models on scenarios that are difficult, dangerous, or costly to collect in real life.
• Provides highly detailed labels, including for complex effects like multi-path reflections, Doppler signatures, and object-specific properties.
• Improves the realism and variety of training data, leading to more robust and accurate AI models.
• Can simulate advanced sensors (higher resolution, less noise) beyond current hardware limitations for research or AI enhancement purposes.
• Applicable to a wide range of sensing modalities (radar, lidar, camera, ultrasound), supporting cross-domain AI training and sensor fusion.
• Facilitates rapid prototyping, validation, and improvement of perception algorithms for emerging fields like autonomous driving and robotics.
• Allows controlled variation of environmental conditions (e.g., weather, sensor configurations), supporting targeted AI evaluation and development.
• Reduces dependency on real-world data collection, which can be expensive, time-consuming, or regulatory restricted.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The present invention relates to a system for detecting, classifying and/or segmenting an object and/or a property of an object with a simulation unit and an artificial intelligence unit, wherein the simulation unit is designed to generate simulated measurement data from a simulated sensor unit relating to a virtual object and/or a property of a virtual object in a virtual environment and to annotate the virtual object, the property of the virtual object and/or the virtual environment and to generate simulation data therefrom, wherein the artificial intelligence unit is designed, on the basis of the simulated measurement data and the simulation data, to detect, classify and/or segment an object and/or a property of an object on the basis of the simulated measurement data or on the basis of other simulated measurement data and/or on the basis of measurement data generated from a real measurement.