Elysia Tools

Browse tools, examples, visual explainers, and topic guides from one clear workspace.

Top Categories

Hubs

Language
ZHDERUPTENFRES

Elysia Tools

Navigation

AI Tools

AI Face Compare (1:1)

· Last updated:2025-12-02

Compare two images, output embedding distance and match boolean by threshold, only compare the first face in each image

Execution

Run this tool

Fill in the form, run the tool, and review the result in one place.

Samples

Examples that match this tool

Related

Continue with connected tools and hubs

Written by the Elysia Tools team
Image355
AI63
Match23
Face8

Tool usage guide

Learn when to use this tool, what it supports, and how real users apply it.

Key facts

Category
AI Tools
Input types
file, number
Output type
json
Sample coverage
4
API ready
Yes

Overview

AI Face Compare (1:1) analyzes two uploaded portraits to locate the first face in each image, returning a JSON response with the embedding distance and a boolean match status based on configurable threshold and confidence settings.

When to use

  • You need to verify whether a photo ID matches a submitted onboarding selfie.
  • You want to measure facial similarity between two pictures taken at different times or under different lighting.
  • You need to test and tune embedding distance thresholds for a face verification process.

How it works

  1. 1Upload the first image into Image A and the second image into Image B (BMP, JPEG, PNG, or GIF formats up to 30 MB each).
  2. 2Set the Match Threshold (default 0.6, range 0.1 to 1.5) to define the maximum distance value considered a match.
  3. 3Adjust the Minimum Confidence (default 0.5, range 0.05 to 0.99) to specify the minimum face detection score required for processing.
  4. 4Run the comparison and inspect the JSON output for the computed embedding distance and boolean match result.

Use cases

  • Identity Verification Audits: Check identity records by comparing user-submitted headshots against reference photos to confirm likeness via JSON output.
  • Threshold Calibration: Test sample image pairs across different Match Threshold and Minimum Confidence values to establish accurate operating thresholds for verification tasks.
  • Duplicate Detection Checks: Evaluate two user profiles suspected of belonging to the same individual by measuring the embedding distance between their profile images.

Compare alternatives

  • Compare one face against a whole gallery of enrolled faces (1:N recognition) instead of a single 1:1 pair.

  • Generate 128D face embeddings for each detected face, without comparing two faces.

  • Detect faces in an image and return bounding box coordinates, without comparing identities.

Who it's for

Best for

  • Comparing two individual portrait images to verify whether they represent the same person.
  • Calculating numerical embedding distance between two faces using custom confidence and match thresholds.
  • Generating machine-readable JSON match results for one-to-one (1:1) facial verification.

Not ideal for

  • Scanning group photos to match multiple faces at once, as the tool only processes the first face in each image.
  • Searching a single face against an entire library or database of images (one-to-many 1:N recognition).

Edge cases & common failures

  • Images containing multiple faces or background bystanders

    The tool only analyzes the first detected face in each file. Pre-crop the images to contain only the single subject of interest before uploading.

  • Face detection fails due to low lighting or steep angles

    If no face is detected, note that the tool already retries automatically at confidences of 0.35 and 0.2 when the default 0.5 setting finds nothing; only lowering Minimum Confidence below 0.2 can help detect difficult faces, or provide a clearer front-facing portrait.

  • Borderline match status near the threshold boundary

    Review the numerical embedding distance in the JSON output and calibrate the Match Threshold between 0.1 and 1.5 to suit your required tolerance for false matches versus false rejections.

Examples

1. Verifying an Onboarding Selfie Against an ID Photo

Identity Verification Reviewer

Background

A platform collects photo identification and a live selfie during new user onboarding.

Problem

The reviewer needs to confirm whether both photos show the same individual despite different framing and lighting.

How to use

Upload the ID photo to Image A and the selfie to Image B, leave Match Threshold at 0.6 and Minimum Confidence at 0.5, then execute the comparison.

Image A: id_photo.png, Image B: selfie.jpg, Match Threshold: 0.6, Minimum Confidence: 0.5

Outcome

The tool returns a JSON object reporting an embedding distance within the threshold and a match status of true.

2. Auditing Suspected Duplicate Profile Photos with Strict Thresholds

Account Security Analyst

Background

An analyst investigates two accounts suspected of being operated by the same individual using different profile avatars.

Problem

The analyst requires a stricter matching threshold to minimize false positives before taking account actions.

How to use

FAQ

Which image file types and size limits are supported?

The tool supports BMP, JPEG, PNG, and GIF files up to 31,457,280 bytes (approximately 30 MB) per image.

How does the Match Threshold parameter work?

Match Threshold sets the cutoff for the embedding distance. A distance at or below this value results in a true match boolean, while a distance above it returns false.

What does the Minimum Confidence parameter do?

It sets the starting detection confidence score (0.05 to 0.99) for registering a face in each image. If no face is found at that level, the tool automatically retries at lower confidences (down to 0.2) before giving up, so only lowering the setting below 0.2 changes which faces can be detected.

What happens if an uploaded photo contains more than one face?

The tool detects and compares only the first face identified in each uploaded image, ignoring any other faces.

What output does the tool provide?

It returns a JSON object containing the calculated embedding distance value and a boolean match determination based on your threshold setting.

Hub

AI Face Analysis: Detection, Comparison, and Recognition Workflows

Separate face detection and alignment from descriptor comparison, gallery recognition, and expression or age-gender estimation, with practical privacy and error boundaries.

Home
Tools3212
Samples593
Visualizations471
Viewers103
Hubs239
Tags333
Images, Audio & Video657

Image, audio, and video processing, conversion, and optimization tools

Math, Date & Finance647

Calculators, numerics, date logic, statistics, and finance tools

Design & Color298

Color, layout, graphics, visual styling, and design helper tools

Text & Writing209

Writing, cleanup, formatting, extraction, and text analysis tools

Developer & Web193

Developer utilities, networking, web debugging, and automation helpers

Conversion & Encoding175

Format, file, archive, unit, and encoding conversion tools

Lead Generation Quiz Launch and Contact Handoff
Memorial Drafts: Family Review and Print Handoff
Private Habit Records: Backup, Restore and Sharing Review
Article to X and LinkedIn Drafts: Fact Review and Handoff
Markdown Interactive Deck Review and Offline Handoff

© 2026 Elysia Tools. All rights reserved.

DocsPricingAboutTerms of ServicePrivacy PolicyCookiesContact
Objective-C

macOS Image Processing Objective-C Samples

macOS Objective-C image processing examples including image reading/saving, image scaling, and format conversion

Go

Web Image Processing Go Samples

Web Go image processing examples including image reading/saving, image resizing, and format conversion

Python

Web Image Processing Python Samples

Web Python image processing examples using PIL/Pillow including reading, saving, resizing, and format conversion

TypeScript

Web Image Processing TypeScript Samples

Web TypeScript image processing examples including image read/save, scaling, and format conversion

Design

Image Adaptive Threshold

Apply adaptive threshold operation that calculates threshold for small regions, suitable for varying lighting conditions

Design

Image Boolean AND

Apply bitwise AND operation between two images to create intersection mask

Design

Image Boolean OR

Apply bitwise OR operation between two images to create union mask

Design

Image Boolean XOR

Apply bitwise XOR operation between two images to create difference mask

Upload the two profile avatars into Image A and Image B, set Match Threshold to 0.40, and check the resulting JSON data.

Image A: avatar_user1.jpeg, Image B: avatar_user2.png, Match Threshold: 0.4, Minimum Confidence: 0.6

Outcome

The tool outputs the exact embedding distance and evaluates the match boolean against the 0.40 limit.

AI Face Recognition (Gallery)
AI Face Descriptors
AI Face Detection