AAO research

AAO AUG222025_01B2203: Computer Vision Evidence Gaps

AAO AUG222025_01B2203: Computer Vision Evidence Gaps

Published material and judging remained accepted, but original contributions and occupational salary comparisons did not establish a third category.

Decision snapshot

Decision detail

Source-backed value

Decision ID

AUG222025_01B2203

Disposition stated in the source

The appeal was dismissed at the initial-criteria stage; the AAO expressly reserved formal final merits despite additional aggregate comments.

Source location

Final order, page 7

Published material and judging were the two credited categories

SCOPS credited published material and judging for the computer scientist. The appeal pursued original contributions and high remuneration. The AAO found neither additional category established, dismissed the appeal, and expressly reserved formal final merits. Its later observations about the aggregate record should not be mistaken for a completed formal merits determination.

Source: AUG222025_01B2203 page 1; AUG222025_01B2203 page 2; AUG222025_01B2203 page 6; AUG222025_01B2203 page 7.

The AAO accepted originality but found major significance insufficiently shown

The AAO recognized original work in object-pose estimation, three-dimensional computer vision, and event-camera processing. Letters described others referring to and comparing the work, but did not adequately explain the claimed major significance. Citation geography and bibliometric assertions also lacked sufficient context; OpenAlex methodology was not documented, and the Google Scholar evidence did not identify self-citations. These were evidentiary shortcomings, not findings that the contributions were unoriginal or that citation data were fabricated.

Source: AUG222025_01B2203 page 2; AUG222025_01B2203 page 3; AUG222025_01B2203 page 4; AUG222025_01B2203 page 5.

Salary comparisons lacked duties needed to connect the occupations

The current offer described a senior applied research scientist position without enough duties to assess its relationship to the wage statistics for computer and information research scientists. Competing offers likewise lacked sufficient duties for comparison. The decision expressly allowed credible prospective compensation evidence in principle; the problem was comparability in this record, rather than a requirement that every offered dollar already have been earned.

Source: AUG222025_01B2203 page 5; AUG222025_01B2203 page 6.

Current framework and related reading

The decisions above are non-precedent assessments of particular records. Historical reasoning is identified as historical; consult current guidance for the present framework.

Primary sources reviewed for this article

Sources and corrections methodology