← USPTO Patent Grants

Visual attribute expansion via multiple machine learning models

Grant US12579220B2 Kind: B2 Mar 17, 2026

Assignee

Microsoft Technology Licensing, LLC

Inventors

Pramod Kumar Sharma, Yijian Xiang, Yiran Li, Paul Pangilinan Del Villar, Liang Du, Robin Abraham, Nilgoon Zarei, Mandar Dilip Dixit

Abstract

A computer implemented method includes receiving an image that includes a type of object, segmenting the object into multiple segments via a trained segmentation machine learning model, and inputting the segments into multiple different attribute extraction models to extract different types of attributes from each of the multiple segments.

CPC Classifications

G06F 18/2431 G06F 18/213 G06N 3/044 G06T 7/11 G06T 2207/20084 G06V 10/46 G06V 10/56 G06V 10/82 G06V 20/60 G06V 10/26

Filing Date

2021-09-21

Application No.

17480869

Claims

20