faceverify

Getting Started

Get up and running with FaceVerify in minutes.

Prerequisites

Python 3.8 or higher
pip package manager
4GB RAM minimum (8GB recommended)

Installation

From PyPI

pip install faceverify-sdk

From PyPI (Pre-release)

pip install faceverify-sdk --pre

From Source

git clone https://github.com/nayandas69/faceverify.git
cd faceverify
pip install -e .

With Development Dependencies

pip install -e ".[dev]"

Verify Installation

python -m faceverify info

Expected output:

FaceVerify v1.0.0rc1
========================================
Python:        3.x.x
Platform:      Windows/Linux/macOS
...

First Verification

Step 1: Prepare Images

Create a folder with test images:

test_images/
  person1_a.jpg   # First photo of person 1
  person1_b.jpg   # Second photo of person 1
  person2.jpg     # Photo of different person

[!TIP] Use clear, front-facing photos with good lighting for best results.

Step 2: Run Verification

Using Python

from faceverify import FaceVerifier

# Initialize
verifier = FaceVerifier()

# Verify same person
result = verifier.verify("test_images/person1_a.jpg", "test_images/person1_b.jpg")
print(f"Same person: {result.verified}")
print(f"Confidence: {result.confidence:.2%}")

# Verify different people
result = verifier.verify("test_images/person1_a.jpg", "test_images/person2.jpg")
print(f"Same person: {result.verified}")
print(f"Confidence: {result.confidence:.2%}")

Using CLI

# Same person
python -m faceverify verify test_images/person1_a.jpg test_images/person1_b.jpg

# Different people
python -m faceverify verify test_images/person1_a.jpg test_images/person2.jpg

Understanding Results

VerificationResult Fields

Field	Description
`verified`	True if faces match
`confidence`	Confidence score (0-1)
`similarity`	Raw similarity score
`threshold`	Threshold used

Example Output

==================================================
  Face Verification Result
==================================================
  Status:      VERIFIED
  Confidence:  92.31%
  Similarity:  0.9231
  Threshold:   0.6500
==================================================

Next Steps

Troubleshooting

No faces detected

[!WARNING] If verification fails with “No face detected”, check:

Image has a clear, visible face

Image is not too dark or blurry

Face is not too small (at least 80x80 pixels)

Incorrect results

[!TIP] If verification gives unexpected results:

Try different detector backends: -d retinaface

Adjust threshold: -t 0.70

Use higher quality images

Import errors

[!CAUTION] If you see ModuleNotFoundError:

Ensure faceverify-sdk is installed: pip install faceverify-sdk

Check Python version: python --version

Activate virtual environment if using one

This site is open source. Improve this page.