Latest News
Introducing the new OFIQ face processing engine (BQAT-CLI) 🚀 Try it now!
BQAT-CLI v1.6.1 released! Join and share your ideas at Discussions 🎉
by Biometix
Overview
BQAT (Biometric Quality Assessment Tool) BQAT is an open-source biometric quality assessment tool for generating and analyzing biometric samples’ quality against international standards as well as customized metrics. BQAT functions by taking an input directory of biometric data and producing raw quality information and an analysis report.
The quality of biometric samples is a key aspect of the performance and efficacy of a biometric system. Whilst there are a variety of tools suitable for each modality, this project is aimed to provide an open-source framework to support all common modalities and allow for expansion as new methods are developed.
graph LR
subgraph title [ ]
core[BQAT-Core] --> face("`__Face__`")
core --> finger("`__Fingerprint__`")
core --> iris("`__Iris__`")
core --> speech("`__Speech__`")
finger --> nfiq([NFIQ2])
iris --> biqt-iris([BIQT])
face --> bqat([Biometix])
face --> ofiq([OFIQ])
face --> biqt-face([BIQT])
speech --> nisqa([NISQA])
end
title --> cli(BQAT-CLI)
title --> api(BQAT-API)
title --> gui(BQAT-GUI)
title --> stateless(BQAT-Stateless)
click finger "https://biometix.github.io/modality/fingerprint.html" _blank
click face "https://biometix.github.io/modality/face.html" _blank
click iris "https://biometix.github.io/modality/iris.html" _blank
click speech "https://biometix.github.io/modality/speech.html" _blank
click cli "https://github.com/Biometix/bqat-cli" _blank
click api "https://github.com/Biometix/bqat-api" _blank
click gui "https://github.com/Biometix/bqat-gui" _blank
click stateless "https://github.com/Biometix/bqat-stateless" _blank
click nfiq "https://github.com/usnistgov/NFIQ2" _blank
click biqt-face "https://github.com/mitre/biqt-face" _blank
click biqt-iris "https://github.com/mitre/biqt-iris" _blank
click bqat "https://github.com/Biometix" _blank
click ofiq "https://www.bsi.bund.de/EN/Themen/Unternehmen-und-Organisationen/Informationen-und-Empfehlungen/Freie-Software/OFIQ/OFIQ_node.html" _blank
click nisqa "https://github.com/gabrielmittag/NISQA" _blank
Modalities
-
The engine for the analysis of fingerprints is based on NIST/NFIQ2 quality features. The quality score links the image quality of optical and ink 500 PPI fingerprints to operational recognition performance.
-
The face image assessment module provides metrics including head pose, smile detection, inter-eye distance, closed eyes, etc.
-
The iris sample assessment module provides various quality attributes, features, and ISO metrics.
-
The speech assessment module provides various quality metrics, including naturalness, colouration, noisiness, etc.
Key features of the project
- Biometrics quality assessment platform for different modalities and processing engines
- A variety of different methods of access, including via CLI, web page and/or web API
- Simple installation via Docker
- Quality report
- Built by an experienced biometric consulting team
- An active open-source community
Supporting libraries
Variation of the Toolbox
BQAT is available in multiple form factors:
-
BQAT in the terminal. A CLI tool for biometrics.
-
BQAT via RESTful API. A self-contained server solution.
-
Stateless version of BQAT API. No data will be stored on the server, could see as serverless function.
-
Simple Web GUI for BQAT API. An easy to use user interface to the services.
Let us know if you have any question on GitHub Discussions!
Roadmap
- Face image analysis module
- Fingerprint image analysis module
- Iris image analysis module
- Speech audio file analysis module
- Result EDA reporting feature
- Image formating feature as preprocess
- Image resizing feature as preprocess
- OFIQ processing engine for face modality
- Duplicates removal feature for output
- Identity removal feature as preprocess
- Document image quality check for face modality
- Contact lens detection feature for face modality
- Face liveness detection feature for face modality
- NFIQ1 as alternative fingerprint processing engine
More to come! 🚀
Please contact us at info@biometix.com if you’d like more information.