ARGIRA v1.4.6: Responsive Accessibility and Uncertainty Communication Interface Update

ARGIRA v1.4.6 is a release of the ARGIRA prototype for automatic alt-text generation and three-layer uncertainty communication. This version extends the user-facing interface and adds a responsive-accessibility fix for keyboard and screen-reader focus on result cards.

It is a new version of ARGIRA v1.4.5 (DOI: 10.5281/zenodo.21959360).

A clearer three-layer uncertainty interface

ARGIRA v1.4.6 introduces a clearer presentation of the information provided around each classification result. The prototype now provides:

• a more explanatory natural-language presentation of the Layer 2 interpretation;
• a more prominent presentation of the instrument's estimate;
• a clearer visual indication when a runner-up classification is present;
• a separate notice when the reliability of an estimate is reduced;
• a reorganized "More information" interface, keeping each expandable section associated with its corresponding control;
• an A+/A- text-size control for adjusting the interface text size.

These interface elements are documented as part of the v1.4.6 public release and are intended to make the prototype's uncertainty information easier to understand and navigate.

Responsive focus visibility

v1.4.6 also fixes a responsive accessibility issue in which a result card receiving keyboard or screen-reader focus could become partially hidden behind the sticky navigation bar.

The fix adapts to the actual height of the navigation area, including when the viewport is narrow or the text-size control changes the navigation layout. The existing focus-management behavior is preserved.

Verification

The responsive fix was verified with an automated Chromium test across both English and Spanish artifacts at viewport widths of 320px, 375px, 414px, and 768px, with additional tests using the text-size control at 130%.

All 12 verification cases passed. The verification protocol waited for scroll position stabilization before measuring the final position, avoiding arbitrary fixed delays in the presence of smooth scrolling.

A verification report accompanying this release documents the methodology, the complete pass/fail matrix, the changes verified, and the SHA-256 checksums of the released HTML artifacts.

Experimental scope and continuity

ARGIRA v1.4.6 does not introduce a new image-classification algorithm, calibration model, or experimental threshold. The underlying heuristic classification, calibration information, runner-up computation, uncertainty communication behavior, and public notice mode remain part of the existing ARGIRA prototype.

The release remains an experimental research artifact and is not presented as a finished product or as validated accessibility technology.

Accessibility verification scope

Manual developer verification with Android TalkBack carried forward from previous releases remains part of the accessibility evidence for the prototype. Apple VoiceOver was not evaluated because an available Apple/VoiceOver testing environment was not available. This remains an untested scope, not a failed validation.

No external user study or participant evaluation was conducted for this release.

Package contents

index_es-1_4_6.html — Spanish-localized prototype artifact.

index_en-1_4_6.html — English prototype artifact.

informe_verificacion_responsive.md — responsive verification report, including methodology, the complete 12-case pass/fail matrix, and SHA-256 checksums of the verified HTML artifacts.

README.md — release overview and usage information.

CHANGELOG.md — version history.

CITATION.cff — machine-readable citation metadata.

LICENSE.txt — license text.

SHA256SUMS.txt — SHA-256 checksums for the release files.

The release provides two independently executable prototype artifacts, in English and Spanish, together with the documentation and verification evidence for this version.

 10.5281/zenodo.21965506 

Note: The DOI shown in the footer of the HTML prototype incorrectly points to the v1.4.4 DOI instead of this version's own DOI (v1.4.6). This does not affect the content, functionality, or verified behavior of this release. It will be corrected in the next version.