US 12,725,198 B2Granted

System and method for providing one or more photorealistic synthetic images of a retail product

AssigneeJio Platforms Limited

InventorsNaveen Kumar Pandey, Balakrishna Pailla, Shailesh Kumar, Abhinav Anand, Dhruval Rana, Ritesh Kumar, JayaSaiKiran Patibandla

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

The present disclosure provides a robust and effective solution to an entity or an organization by enabling the entity to implement a system for 3D vision, and synthetic data generation of a retail product model. Further, the present disclosure aims at generating high-quality, photorealistic, large-scale domain-specific synthetic data. The system is equipped with synthetic simulation, digital twin creation, and synthetic data generation with automated annotations. Additionally, the system allows the training of artificial intelligence to reduce manual efforts and to facilitate the automation and personalization.

Claims10

Dependent claims

A system for providing one or more photorealistic synthetic images of a retail product, said system comprising:

one or more processors operatively coupled to one or more computing devices, the one or more processors coupled with a memory, wherein said memory stores instructions which when executed by the one or more processors causes the one or more processors to:
receive one or more input parameters from the one or more computing devices using a configuration template, wherein the one or more computing devices are associated with one or more users, and are connected to the one or more processors through a network,
wherein the one or more input parameters are indicative of one or more images of the retail product provided by the one or more users through the one or more computing devices;
extract a first set of attributes from the one or more input parameters, wherein the first set of attributes are indicative of one or more configuration parameters of the retail product;
extract a second set of attributes based on the first set of attributes, wherein the second set of attributes are indicative of one or more retail product meshes based on the one or more configuration parameters of the retail product;
extract a third set of attributes based on the second set of attributes, wherein the third set of attributes are indicative of one or more digital twins based on the one or more retail product meshes of the retail product;
based on the first set of attributes, the second set of attributes, and the third set of attributes, generate a rendered model through an artificial intelligence (AI) engine, wherein the Al engine is configured to use one or more techniques;
and
generate the one or more photorealistic synthetic images and one or more annotations of the retail product based on the rendered model, wherein the one or more techniques used by the AI engine comprise one or more alpha-blending based techniques that perform domain randomization of the one or more digital twins to generate the one or more photorealistic synthetic images and the one or more annotations of the retail product, and wherein the AI engine is configured to generate one or more three-dimensional (3D) realistic scenes using the one or more alpha-blending based techniques, and simulate the one or more 3D realistic scenes to generate the one or more photorealistic synthetic images of the retail product.

6 more elements

A method for providing one or more photorealistic synthetic images of a retail product, said method comprising:

receiving, by one or more processors, one or more input parameters from one or more computing devices, wherein the one or more computing devices are associated with one or more users, and are connected to the one or more processors through a network,
wherein the one or more input parameters are indicative of one or more images of the retail product provided by the one or more users through the one or more computing devices;
extracting, by the one or more processors, a first set of attributes from the one or more input parameters, wherein the first set of attributes are indicative of one or more configuration parameters of the retail product;
extracting, by the one or more processors, a second set of attributes based on the first set of attributes, wherein the second set of attributes are indicative of one or more retail product meshes based on the one or more configuration parameters of the retail product;
extracting, by the one or more processors, a third set of attributes based on the second set of attributes, wherein the third set of attributes are indicative of one more digital twins based on the one or more retail product meshes of the retail product;
generating, by the one or more processors, based on the first set of attributes, the second set of attributes, and the third set of attributes, a rendered model through an artificial intelligence (AI) engine;
generating, by the one or more processors, the one or more photorealistic synthetic images and one or more annotations of the retail product based on the rendered model;
generating, by the one or more processors through the AI engine, the one or more photorealistic synthetic images associated with the one or more annotations of the retail product using one or more alpha-blending based techniques that perform domain randomization of the one or more digital twins;
and
generating, by the AI engine, one or more three-dimensional (3D) realistic scenes using the one or more alpha-blending based techniques, and generating the one or more photorealistic synthetic images of the retail product by simulating the one or more 3D realistic scenes.

7 more elements

A user equipment for providing one or more photorealistic synthetic images of a retail product, said UE comprising:

one or more processors communicatively coupled to one or more processors comprised in a system, the one or more processors coupled with a memory, wherein said memory stores instructions which when executed by the one or more processors causes the UE to:
transmit one or more images of the retail product to the one or more processors using a configuration template,
wherein the one or more processors are configured to:
receive one or more input parameters from the UE, wherein the UE is associated with one or more users and is connected to the one or more processors through a network,
wherein the one or more input parameters are indicative of the one or more images of the retail product provided;
extract a first set of attributes from the one or more input parameters, wherein the first set of attributes are indicative of one or more configuration parameters of the retail product;
extract a second set of attributes based on the first set of attributes, wherein the second set of attributes are indicative of one or more retail product meshes based on the one or more configuration parameters of the retail product;
extract a third set of attributes based on the second set of attributes, wherein the third set of attributes are indicative of one or more digital twins based on the one or more retail product meshes of the retail product;
based on the first set of attributes, the second set of attributes, and the third set of attributes, generate a rendered model through an artificial intelligence (AI) engine, wherein the AI engine is configured to use one or more techniques;
and
generate the one or more photorealistic synthetic images and one or more annotations of the retail product based on the rendered model, wherein the one or more techniques used by the AI engine comprise one or more alpha-blending based techniques that perform domain randomization of the one or more digital twins to generate the one or more photorealistic synthetic images and the one or more annotations of the retail product, and wherein the AI engine is configured to generate one or more three-dimensional (3D) realistic scenes using the one or more alpha-blending based techniques, and simulate the one or more 3D realistic scenes to generate the one or more photorealistic synthetic images of the retail product.

8 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/722,384Filed 26 Dec 2022
Priority: IN 20212106097227 Dec 2021
Published: US 2025/0139689 A11 May 2025
Granted: 1 Sep 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

G06Q 30/0643
Utilising user interfaces specially adapted for shopping: graphically representing goods, e.g. 3D product representation
G06T 17/00
Three-dimensional [3d] modelling for computer graphics
G06T 7/60
Image analysis: analysis of geometric attributes
G06T 15/04
Three-dimensional [3D] image rendering: texture mapping