Using image augmentation with simulated objects for training machine learning models

A technology for simulating objects and images, applied in the field of training machine learning models, can solve problems such as difficult to generate a practical, stable and reliable automatic driving system, and achieve the effect of improving decision-making results, increasing accuracy, and reducing high-precision time

Pending Publication Date: 2021-10-12
NVIDIA CORP
View PDF1 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

As a result, it is increasingly difficult to generate practical, stable and r...

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Using image augmentation with simulated objects for training machine learning models
  • Using image augmentation with simulated objects for training machine learning models
  • Using image augmentation with simulated objects for training machine learning models

Examples

Experimental program
Comparison scheme
Effect test

example 204

[0070] Figure 8 Is a flowchart of a method 800 for training a machine learning model using real world images augmented with simulated objects to detect objects in the images according to some embodiments of the present disclosure. At block B802, method 800 includes generating a first image of a virtual instance of an object from a virtual sensor perspective of a virtual vehicle in a virtual environment. For example, from the perspective of virtual sensors installed on a virtual vehicle in a simulated environment, a game-based engine such as Figure 2A Virtual instances 204, 206 and 208 of dead animals are shown.

[0071] At block B804, the method 800 includes determining a boundary shape corresponding to the virtual instance of the object. For example, a boundary shape (such as boundary shape 304 ) corresponding to a virtual instance (such as virtual instance 206 ) can be created using segmentation mask 214 associated with virtual instance 206 . In some embodiments, a boun...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

Using image augmentation with simulated objects for training machine learning models are disclosed. In various examples, systems and methods are disclosed that preserve rich, detail-centric information from a real-world image by augmenting the real-world image with simulated objects to train a machine learning model to detect objects in an input image. The machine learning model may be trained, in deployment, to detect objects and determine bounding shapes to encapsulate detected objects. The machine learning model may further be trained to determine the type of road object encountered, calculate hazard ratings, and calculate confidence percentages. In deployment, detection of a road object, determination of a corresponding bounding shape, identification of a road object type, and/or calculation of a hazard rating by the machine learning model may be used as an aid for determining next steps regarding the surrounding environment, e.g., navigating around the road debris, driving over the road debris, or coming to a complete stop, in a variety of autonomous machine applications.

Description

[0001] Cross References to Related Applications [0002] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 003,879, filed April 1, 2020, which is hereby incorporated by reference in its entirety. Background technique [0003] Autonomous and semi-autonomous vehicles utilize machine learning (e.g., deep neural networks (DNN)) to analyze the road surface as the vehicle travels in order to guide the vehicle's position relative to road boundaries, lanes, road debris, barriers, road signs, etc. . For example, DNNs can be used to detect road debris (e.g., animals, cones, building materials) in oncoming parts of the road while the ego vehicle is driving, which may lead to adjustments to the ego vehicle's position (e.g., manipulating to avoid passing the traffic cone in the middle of the road). However, training DNNs to accurately detect objects on the road requires a large amount of training data, computational power, and human time and effort. Furt...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06N20/00G06N3/04G06N3/08
CPCG06N20/00G06N3/08G06N3/045B60W60/001G06N3/04B60W2554/80B60W2420/42B60W2554/4029B60W50/06
Inventor 崔泰银郝鹏飞林晓琳M·帕克
Owner NVIDIA CORP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap