Vision-language models could spot mobility barriers with expert guidance
Researchers tested whether large vision-language models could spot mobility barriers in street-level imagery by combining such models with guidance informed by accessibility standards and expert-created assessment tools.
At a university campus, researchers linked 407 locations from street-level imagery with GPS data tracking where wheelchair users spent more time, treating that dwell time as an indicator of possible mobility obstacles.
When enhanced with expert guidance, the strongest model's correlation between accessibility ratings and dwell time improved from r=-0.40 to r=-0.54.
The models reliably identified visible features like curb ramps and crosswalks as linked to higher accessibility scores, but showed limited ability to detect subtle surface problems, temporary blockages, and barriers that depend on viewing angle.
Sources behind this report
- arXiv
Research · abstract · CC0 1.0
Research manuscript in arXiv. Peer-review status is not recorded here; consult the linked version and publication.
Source published
Read by the newsroom
Report history
No earlier revision dates are retained here. This does not establish that the report has never changed; read any dated updates and corrections.
This brief was made by AI and checked against its cited public sources. How the newsroom works