Invention Title:

SYSTEMS AND METHODS FOR PROVIDING A DIGITAL HUMAN IN A VIRTUAL ENVIRONMENT

Publication number:

US20240320519

Publication date:

2024-09-26

Section:

Physics

Class:

G06N5/022

Inventors:

Chellappan MURUGAPPAN Chennai, India

Ashok VIRA Mumbai, India

Bolaka Mukherjee Howrah, India

Raghvendra KOTALWAR Pimple Saudagar, India

Ankur TANDON Ghaziabad, India

Amit MISRA Uttar Pradesh, India

Ashwin Joseph D SILVA Mumbai, India

Anandu KOIKKALETHU BALACHANDRAN Pathanamthitta, India

Assignee:

ACCENTURE GLOBAL SOLUTIONS LIMITED Dublin 4, Ireland

Applicant:

ACCENTURE GLOBAL SOLUTIONS LIMITED Dublin 4, Ireland

Drawings (4 of 16)

Drawing 01 for SYSTEMS AND METHODS FOR PROVIDING A DIGITAL HUMAN IN A VIRTUAL ENVIRONMENT

Drawing 02 for SYSTEMS AND METHODS FOR PROVIDING A DIGITAL HUMAN IN A VIRTUAL ENVIRONMENT

Drawing 03 for SYSTEMS AND METHODS FOR PROVIDING A DIGITAL HUMAN IN A VIRTUAL ENVIRONMENT

Drawing 04 for SYSTEMS AND METHODS FOR PROVIDING A DIGITAL HUMAN IN A VIRTUAL ENVIRONMENT

Smart overview of the Invention

Innovative systems and methods are designed to create a digital human within a virtual environment. Users can select from various personas for their digital human, allowing for a personalized interaction experience. The system processes user inputs in real-time, identifying specific parameters related to the input to tailor responses effectively.

Functionality of the System

The system comprises a processor and memory that work together to facilitate interactions. Upon receiving a user's persona selection and input, the processor identifies parameters like context and emotional tone. It then retrieves relevant response information from a knowledge database and generates attributes that define how the digital human should respond, incorporating both audio and visual elements.

Response Generation Process

To determine appropriate responses, the processor employs advanced techniques such as generating embeddings from user inputs. These embeddings help in selecting relevant information based on similarity parameters, ensuring that responses are contextually relevant. Depending on the analysis outcome, different machine learning models may be used to refine the response further.

Attributes and Personalization

The system generates a set of attributes that influence how the digital human communicates. This includes audio features like voice tone and emotion, alongside visual attributes such as facial expressions and gestures. By aggregating these attributes with the response information, the system can produce a personalized interaction that feels more human-like.

Implementation in Various Platforms

The technology is adaptable across multiple digital platforms, including messaging services and applications. The ability to capture and analyze user behavior allows for continuous improvement of persona profiles, enhancing future interactions. Overall, this approach aims to bridge the gap between human communication and machine responses, fostering more engaging user experiences.