Voice Fingerprint Authentication With ResCNN for Remote Exam Integrity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Challenges exist in verifying the identity of remote examination candidates securely and reliably in online examination settings, risking examination integrity and public confidence.
Innovation Solution
A method and system for training an AI neural network using a residual convolutional neural network (ResCNN) to authenticate voice fingerprints, incorporating a softmax function and liveness detection, with threshold matching and liveness checks, to ensure accurate candidate verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional candidate verification methods are used in remote proctored examinations, then the system is simpler to implement, but the security and reliability of candidate identity verification deteriorates
Solution Approach 1:
The patent replaces traditional mechanical/manual verification methods with an AI-based voice fingerprint recognition system. The ResCNN neural network automatically extracts and compares voice features, substituting human judgment with automated biometric authentication. This substitution dramatically improves verification reliability while the modular architecture manages system complexity through standardized processing pipelines.
Solution Approach 2:
The patent introduces voice fingerprints as an intermediary element between the candidate and the verification system. Instead of directly verifying candidate identity through complex document checks or human inspection, the system uses voice fingerprints as a mediating biometric marker that can be automatically extracted, stored, and compared. This intermediary approach simplifies the verification process while maintaining high security standards.
2Measurement precision
If advanced AI neural network training is implemented for voice authentication, then the measurement precision of candidate verification improves, but the loss of time for data processing increases
Solution Approach 1:
The patent performs voice fingerprint extraction and neural network training in advance, before actual examination verification is needed. The ResCNN model is pre-trained on extensive voice datasets during system setup, creating a ready-to-use authentication engine. This preliminary action allows the system to achieve high measurement precision through thorough training while minimizing time loss during actual candidate verification, as the heavy computational work is completed beforehand.
Solution Approach 2:
The patent implements dynamic processing where the neural network adapts its processing depth based on verification needs. During actual examination, the system uses the pre-trained model for rapid voice fingerprint comparison rather than re-training. The system dynamically adjusts between comprehensive training mode (for model improvement) and quick verification mode (for candidate authentication), optimizing the balance between precision and time efficiency.
Data Source
AI summary
A method and system of training an artificial intelligence (AI) neural network in authenticating a voice fingerprint. The method includes extracting data from a training dataset of voice fingerprints with an AI neural network that includes a residual convolutional neural network (ResCNN), generating processed voice fingerprint data from the extracted data based at least in part on a softmax function implemented in accordance with a softmax layer of the ResCNN, preparing a training dataset and a validation dataset of voice fingerprints based on the processed voice fingerprint data, training the AI neural network based on the training dataset and validating the trained AI neural network based at least in part on the validation dataset.