ageitgey/face_recognition
56.5
Adequate · 14 November 2026
1.8k
lines of production code
Python
primary language
7
measurements over time
What this system is
Features
Major CLI and API overhaul with new face detection tool
The command-line interface has been split into two distinct tools: face\_recognition\_cli.py for identifying known people and face\_detection\_cli.py for detecting faces. The API has been significantly expanded to support the CNN-based face detection model (in addition to the existing HOG method) and batch processing of images for improved performance. Additionally, the library now exposes the face\_distance function and batch\_face\_locations for advanced usage, while removing the old single-image processing approach in favor of more flexible, batched operations.
_face\recognition · high confidence
New Dockerfiles for CPU, GPU, and Jupyter environments
Added new Dockerfiles to build and run the face recognition library in various environments: a standard CPU-based image, a GPU-accelerated image, and Jupyter-based images for both CPU and GPU. The CPU image installs dlib v19.21 and face\_recognition from source, while the GPU image additionally compiles dlib with CUDA support. Documentation in the README explains prerequisites for GPU usage and provides an example Dockerfile for custom applications.
docker · high confidence
New Jupyter Notebook example for real-time face tracking in video
Added a new Jupyter Notebook example demonstrating how to detect and track faces in video in real-time using OpenCV and matplotlib. The notebook includes code for importing necessary libraries and displaying the results interactively.
_examples/ipynb\examples · high confidence
New face recognition examples and tools
Added several new example scripts to the examples directory, including face recognition using KNN and SVM classifiers, blink detection, face blurring, and performance benchmarking. Also added a face distance utility and a web service example.
examples · high confidence
Behavioural changes
Major release 1.4.0: Python 2 removal, CLI enhancements, and Docker support
This release drops support for Python 2.x and adds support for Python 3.8 and 3.9. The CLI is enhanced with a new \face\_detection\ command alongside the existing \face\_recognition\ tool, and the \--upsample\ parameter is added to the face recognition CLI. The project also introduces official Docker support with \Dockerfile\ and \Dockerfile.gpu\ for containerized deployment, alongside a \docker-compose.yml\ for easy setup. Additionally, the library's dependencies are updated to require \face\_recognition\_models\ and \Pillow\, while removing the \scikit-image\ dependency.
(repo-wide) · high confidence
Test coverage
Expanded test coverage for face recognition API
Added and updated tests for the face recognition module, including validation for 32-bit images, CNN-based face detection, batched face location detection, and face distance calculations. The test suite now also covers the small face model for landmarks and verifies the face\_distance function returns the correct numpy array type.
tests · high confidence
Dependencies
Update project dependencies and add build system configuration
The project now includes a pyproject.toml file specifying setuptools and wheel as build-system requirements. The requirements\_dev.txt file has been updated to upgrade pip from 8.1.2 to 21.1, cryptography from 1.7 to 3.3.2, and PyYAML to a newer version, while also adding face\_recognition\_models, Click, dlib, numpy, and scipy to the development dependencies.
(dependencies) · medium confidence
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
How this codebase got here
Score
- CAI 56 → 56 (+0.4)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.4) — scores are not directly comparable.
Lenses
- Code Health 99 → 91 (-8.4)
- Architecture 96 → 96 (+0.0)
- Maturity 48 → 61 (+12.2)
- Readiness 44 → 41 (-2.2)
- Security 76 → 68 (-8.4)
Resolved (4)
- Duplicate utility function. The exact same signature 'image_files_in_folder(folder)' exists in both the detection CLI and recognition CLI modules. This indicates code duplication rather than a shared utility module.
- Inconsistent naming and signature for result printing. One prints a 'location' (detection context), the other prints 'name' and 'distance' (recognition context), but both are named 'print_result'. This suggests a shared utility or interface that has diverged, causing confusion about whether 'print_result' is a generic formatter or domain-specific.
- Inconsistent naming for similar high-level operations. Both modules have a 'test_image' method, but they serve different purposes (detection vs. recognition) and have vastly different signatures. While the names are consistent with their domains, the parallel naming might mislead users into thinking they are interchangeable or part of a unified interface.
- No ADRs found
New (7)
- Edited copy of a member (1 corresponding lines) (face_recognition/face_detection_cli.py)
- Edited copy of a member (11 corresponding lines) (face_recognition/face_detection_cli.py)
- Edited copy of a member (19 corresponding lines) (face_recognition/face_detection_cli.py)
- Edited copy of a member (28 corresponding lines) (examples/face_recognition_knn.py)
- Edited copy of a member (34 corresponding lines) (examples/face_recognition_knn.py)
- Edited copy of a member (48 corresponding lines) (examples/web_service_example.py)
- No ADRs found (setup.py)
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
Survey your own repository
ageitgey/face_recognition was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.
About this page
- The score is its most recent published measurement, taken on 14 November 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
- Measured at commit 9f3061aaeed9a8756d2c970f5dfe066617a8281d — the exact code this score is about.
- Scored under rubric-2026.11.4 — the same rubric and the same method as every other entry in this index.
- Measured by Canine Code Analyzer (cca.canine.dev) using cca measure-2026.11.65+preprod-36d4cd83012f.