Unlock instant, AI-driven research and patent intelligence for your innovation.

Fine-grained object recognition in robotic systems

A robot system and object technology, applied in the field of robot systems, can solve problems such as difficult detection and recognition of objects

Active Publication Date: 2019-04-16
HUAWEI TECH CO LTD
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Unique objects are more difficult to detect and identify

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
  • Fine-grained object recognition in robotic systems
  • Fine-grained object recognition in robotic systems
  • Fine-grained object recognition in robotic systems

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] Subsequent descriptions of the method and robotic system hardware modules relate to implementation in a domestic service robot. This is for illustration purposes only, the fine-grained object recognition can be implemented in other machines (eg robotic systems).

[0032] Subsequent discussions refer to objects as object classes (ie, class-level objects) and object instances (instance-level objects). Object classes or class-level objects can be defined as a type of object. For example, cups, phones, shoes, and bottles are considered classes of objects. An instance of an object or instance-level object refers to the make, type, or even model of that particular object class. In the example of a phone object, the instance of the phone would be a cell phone relative to a home phone, a specific brand of phone, or even a specific model of a certain brand of phone (ie, iPhone 4 vs. iPhone 6).

[0033] The disclosed embodiments provide by combining deep neural networks, deep ...

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

A method for fine-grained object recognition in a robotic system is disclosed that includes obtaining an image of an object from an imaging device. Based on the image, a deep category-level detectionneural network is used to detect pre-defined categories of objects. A feature map is generated for each pre-defined category of object detected by the deep category-level detection neural network. Embedded features are generated, based on the feature map, using a deep instance-level detection neural network corresponding to the pre-defined category of the object, wherein each pre-defined categoryof an object comprises a corresponding different instance-level detection neural network. An instance-level of the object is determined based on classification of the embedded features.

Description

[0001] Related Applications Cross Application [0002] This application claims priority to a prior application to U.S. Nonprovisional Patent Application No. 15 / 449,541, filed March 3, 2017, entitled "Fine-Grained Object Recognition in Robotic Systems," the content of which Incorporated herein by reference. Contents of the invention [0003] The present invention relates to robotic systems, and more particularly to fine-grained object recognition in deep detection embedded networks for home service robots. Background technique [0004] An important function of home service robots is the ability to recognize small indoor objects such as cups, plates, controllers, mobile phones, keys, and bottles. However, due to the personalization of some objects and the resulting fine-grained recognition problem, it is difficult to recognize small objects. [0005] Objects that need to be recognized by a home service robot in a particular home may be personalized by the user, with unique c...

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): G06T7/00G06K9/00G06V10/141G06V10/25G06V10/764G06V10/772G06V10/774
CPCH04N7/185B25J9/1697G06V20/10G06V10/141G06V10/25G06V10/82G06V10/764G06V10/7715G06V10/772G06V10/774G06F18/2413G06F18/2411G06F18/28G06F18/214G06F18/21355
Inventor 蒋伟王炜
Owner HUAWEI TECH CO LTD