Facial Encoding Verification for Video Identity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video conferencing systems face challenges in efficiently verifying the identity of registered users at scale without overburdening computing resources, as they typically require substantial bandwidth and memory to transmit live images or video for verification.
Innovation Solution
The method involves storing facial encodings of registered users, which are derived from images, and comparing these encodings in real-time during video calls to determine identity, reducing the need for transmitting large video data and thus conserving bandwidth and memory resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If live images or video are transmitted for identity verification, then verification accuracy is improved, but bandwidth and memory resources are excessively consumed
Solution Approach 1:
The patent extracts only the essential facial features from complete video frames to create facial encodings. Instead of transmitting entire video streams for verification, the system extracts and transmits only the critical facial encoding data, dramatically reducing bandwidth and memory requirements while maintaining verification accuracy
Solution Approach 2:
The system creates simplified representations (facial encodings) that copy only the essential identifying features from original facial images. These encoded representations serve as lightweight substitutes for transmitting full-resolution video or images, achieving verification purposes with minimal data transmission
2Speed
If facial encodings are compared in real-time during video calls, then identity verification speed is improved, but computing resources are consumed
Solution Approach 1:
The system performs preliminary action by pre-computing and storing facial encodings during user registration, before the actual verification occurs. During real-time video calls, the system only needs to compare incoming facial encodings against pre-stored templates, significantly reducing the computational burden during high-speed verification operations
Solution Approach 2:
The patent transforms complex facial image data into simplified numerical parameter representations (facial encodings). This parameter transformation converts high-dimensional image data into compact numerical vectors that can be rapidly compared using efficient mathematical operations, balancing verification speed with computing resource consumption
3Productivity
If facial encodings are stored in a database, then verification scalability is improved, but data storage requirements increase
Solution Approach 1:
The system extracts only the essential identifying features from complete facial images to create compact facial encodings. By storing only these extracted essential features rather than full images or video streams, the system achieves high verification scalability while maintaining minimal data storage requirements
Solution Approach 2:
The patent transforms facial images into simplified numerical parameter representations that occupy minimal storage space. This parameter transformation enables the system to store and manage large numbers of user profiles efficiently, directly improving verification scalability without proportionally increasing storage requirements
Data Source
AI summary
Systems and methods for verifying an identity of a party to a video call are disclosed. An example method may be performed by one or more processors of a verification system and include storing first facial encodings, each of the first facial encodings associated with one or more images of a corresponding registered user of a plurality of registered users associated with the verification system, receiving, during a video call associated with a specified registered user of the plurality of registered users, a second facial encoding associated with one or more images of a person participating in the video call, the second facial encoding received from a device with which the person uses to participate in the video call, identifying, among the first facial encodings, a first facial encoding corresponding to the specified registered user, and determining whether the person participating in the video call is the specified registered user.