Collect only what is needed
Define the sizing purpose, obtain permission, explain local or secured processing, and avoid using facial measurements for identity recognition.
Protective mask sizing · Facial geometry · Computer vision
Fit should be measured.
MaskFitter is exploring a smartphone-based way to measure facial geometry and identify candidate protective-mask sizes before physical fit testing. The starting point is a documented U.S. defense research problem, not a claim that MaskFitter is a government contractor.
- CaptureMeasure relevant facial geometry
- CompareScreen against mask dimensions
- ValidateTest predicted fit in the real world
Original problem / U.S. defense research
A documented need to size protective masks using a smartphone.
In 2016, the Department of Defense Chemical and Biological Defense (CBD) SBIR program published topic CBD161-005, Smartphone Application for Mask Sizing and Projecting Quantitative Fit. The proposed approach used photographs and a reconstructed 3D headform to estimate facial dimensions, recommend respiratory protective mask sizing, and explore quantitative fit prediction.
The official SBIR records show awards to Technology Solutions Experts, Inc. for Phase I in 2016, Phase II in 2017, and sequential Phase II work in 2019, together totaling approximately $2.13 million. These are historical, completed award records to another company, not an open RFP or an award to MaskFitter.
DoD SBIR Phase I, 2016 (CBD161-005) ↗ · Phase II, 2017 ↗ · Sequential Phase II, 2019 ↗
Updated validation need / 2026
The federal problem has not gone away.
On May 14, 2026, CDC/NIOSH published App Validation Study, solicitation 75D30126Q78873, seeking validation of a NIOSH-developed beta smartphone/tablet application. The app measures faces, classifies people into NIOSH respirator-sizing panels, and predicts respirator models with a likelihood of fitting. The solicitation covered comparison with calipers, human participant testing, usability, and supplementary respirator fit tests.
That opportunity's offers were due June 12, 2026 and the notice is now inactive. It is cited as independent federal research demand, not an available MaskFitter contract, collaboration, or ownership claim.
CDC/NIOSH RFQ 75D30126Q78873 on SAM.gov ↗ · NIOSH's published smart-device respirator-fit research target ↗
MaskFitter's proposed first application
From facial measurements to a mask shortlist.
The intended initial path is protective mask sizing and pre-fit screening. These are development hypotheses, not a functioning, government-qualified product.
Extract face geometry
Use a compatible phone and guided capture to estimate anthropometric landmarks, face width and length, facial contours, and relevant mask-contact regions.
Screen mask candidates
Compare measurements with validated manufacturer and model-specific sizing data, when available, to produce a transparent shortlist rather than a safety guarantee.
Test the prediction
Compare recommendations with physical measurements and accepted qualitative or quantitative respirator fit tests. Track errors across different face geometries and capture conditions.
The first prototype would need evidence of repeatable face measurements, reliable candidate rankings, bias and usability testing, and agreed handling of sensitive geometry data before any real-world performance claim.
Beyond the initial use case
One measurement approach. Several possible fitment markets.
Protective facepiece sizing provides the initial problem definition, but the underlying geometry-to-equipment comparison could be evaluated in industrial respiratory protection, healthcare and CPAP mask selection, and other head- or face-worn equipment such as goggles, helmets, or specialized protective gear.
Each application needs its own dimensional data, validation subjects, performance endpoints, privacy design, commercial agreements, and regulatory or safety review. A successful protective-mask sizing experiment would not automatically validate a CPAP mask, safety respirator, helmet, or medical claim.
Safety, privacy and evidence
Screening is not certification.
OSHA requires fit testing of tight-fitting workplace respirators before initial use, when a different facepiece is used, and at least annually thereafter under 29 CFR 1910.134 ↗. A smartphone suggestion or projected fit factor does not replace required qualitative or quantitative fit testing.
Face geometry can be sensitive information. Any future implementation must define explicit consent, capture quality, minimum retention, access control, human oversight, and limitations on secondary use. Geometry-first is a design intention, not proof of anonymous or risk-free processing.
MaskFitter is a concept-stage initiative of Curate Holdings, informed by publicly documented government research. MaskFitter has no claimed DoD, CDC, NIOSH, or awardee affiliation; no protective performance has been demonstrated or certified. It is not currently an operating medical or protective-fit service.
Interested in research validation, equipment sizing data, or applied fitment partnerships? Discuss the research direction ↗