Facial Recognition as Neural-Network Authentication
Which of the following is the best example of an AI model that is trained to identify multiple points from input using a neural network to provide output for authentication?
Community Insight
The question tests whether you can distinguish an AI/ML biometric authentication model from non-AI security constructs; the trap is picking a security term like OAuth or an encryption key because it sounds authentication-related.
This CompTIA CySA+ (CY0-001) question asks which option is an AI model that uses a neural network to map multiple points from an input and output a result for authentication. Facial recognition is the correct answer, because it is a deep-learning biometric that detects and compares numerous facial landmark points (eyes, nose, jawline) to verify identity, while the other options are cryptographic secrets, an authorization protocol, and a computer-vision annotation technique.
The most common wrong pick is Open Authorization (OAuth), which learners select because the name implies authentication—but OAuth is a delegated authorization framework that exchanges tokens, not a neural-network model that analyzes input points. Bounding box and encryption key are also non-neural-network concepts.
Community Discussion (2 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
Facial recognition is a deep neural-network model (typically a CNN) that first performs face detection, then extracts dozens or hundreds of landmark points and a numeric face embedding from the input image, and finally compares that embedding against a stored template to authenticate the user. This matches every element of the stem: an AI model, neural-network based, identifying multiple points from input, and producing output used for authentication. It is the classic example of AI applied to identity and access management, and vendors such as NIST treat biometric matching under digital identity assurance as an analytic process. Both Kristen8730 and Mu2 converged on this answer, with Kristen8730 accurately describing the mechanism: 'Facial recognition because it scans face multiple points and provides authentication.'Why the Other Options Are Wrong
Open Authorization (OAuth) is a token-based delegation protocol; it grants third-party applications limited access without exposing credentials, but it involves no neural network and no point-mapping of biometric input—it authorizes, it does not authenticate. An encryption key is a symmetric or asymmetric cryptographic secret used to transform data, not an AI model at all. A bounding box is a rectangle annotation produced by object-detection models (used in training data or image localization) to frame an object's location; it identifies where something is in an image, not who the person is, and it is not used for authentication. Only facial recognition combines neural-network inference with identity verification.Community Comment Notes
The two substantive comments both chose A and were correctly reasoned. Kristen8730's comment is the most precise: 'Facial recognition because it scans face multiple points and provides authentication,' which maps directly onto the stem's 'identify multiple points' and 'provide output for authentication.' Mu2's comment that 'only facial recognition uses LLM (neural networks)' reaches the right answer but loosely labels the technique a large language model—facial recognition is actually powered by convolutional neural networks, not LLMs—so the conclusion is right even though the terminology is imprecise.Official Reference
Exam Strategy
On CySA+ questions, identify the core action in the stem—here 'neural network,' 'multiple points,' and 'authentication'—and eliminate options that are protocols, cryptographic material, or model outputs rather than models themselves. Remember that biometric AI (facial recognition, fingerprint, voice) is the go-to authentication example in the AI objectives.
Frequently Asked Questions
Why is OAuth not the correct answer for a neural-network authentication model?
OAuth is a token-based delegated authorization protocol that lets apps access resources on a user's behalf. It exchanges tokens and grants permissions but performs no neural-network analysis of biometric input points.
What is a bounding box and why doesn't it authenticate anyone?
A bounding box is the rectangle an object-detection model draws to localize an object within an image, often used in training data labeling. It describes where something is, not whose identity it is, so it cannot support authentication.