MétaCan
Menu
Back to cohort
Record W1538636281

Producing informative text alternatives for images

2012· article· en· W1538636281 on OpenAlexaff
Lisa Tang

Bibliographic record

VenueUniversity Library - University of Saskatchewan (University of Saskatchewan) · 2012
Typearticle
Languageen
FieldArts and Humanities
TopicSubtitles and Audiovisual Media
Canadian institutionsUniversity of Saskatchewan
Fundersnot available
KeywordsInformation retrievalComputer scienceClosed captioningSearch engine indexingImage (mathematics)Representation (politics)Web pageInformation needsWorld Wide WebArtificial intelligence
DOInot available

Abstract

fetched live from OpenAlex

A picture may be worth a thousand words but what might those words be? How do we go about finding those words? Images are often used to convey information, supplement textual content, and/or add visual appeal to documents. Unless the user can see the image and properly interpret it, the user may not receive the same information. While containers exist for providing text alternatives in various types of electronic documents (including Web pages), they are rarely used. When they are used, the text alternatives are not informative. While guidance currently exists regarding which containers to use in order to provide text alternatives, there is little guidance available regarding what information to include in these containers and how to compose text alternatives. The purpose of this work is to establish a procedure for identifying the information being communicated within an image and provide guidance on how to produce informative text alternatives. Based on related information in the areas of Web accessibility, library cataloguing, captioning and audio description, image retrieval and indexing, art description, and tactile representation, important information communicated by an image were identified and a procedure for producing informative text alternatives using that information was developed. Studies were conducted to determine the effectiveness of the procedure to identify important information about an image. Study 1 determined the information identified about an image when the procedure was not available. It also suggested reasons why people would opt to not provide a text alternative for an image. Study 2 determined the information that people would identify when they were given the procedure and a set of questions to help identify information about an image. Study 3 determined the information people identified when they were required to consider all of the different types of information that may be important in an image. The results from these three studies were compared to determine the effectiveness of the procedure to identify important information about the image. Study 4 presented the information identified in the previous three studies to sighted and visually impaired users to evaluate the importance of such information. This study determined the quality of the information identified in the first three studies and the ability of the procedure to identify important information for a wide set of images. The results of these studies showed that the procedure was effective in identifying a greater amount of important information than without the procedure. Additional guidance was also identified to further help people create informative and useful text alternatives. The studies also showed that the procedure could be applied by different user groups to a wide range of images. The procedure was submitted to the International Standards Organization to become a technical specification, which will be available to people around the world.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Insufficient payload (model declined to judge)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Qualitative · Consensus signal: Qualitative
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.094
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0010.000
Science and technology studies0.0010.001
Scholarly communication0.0000.005
Open science0.0010.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0040.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.017
GPT teacher head0.182
Teacher spread0.166 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

Study designQualitative
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations4
Published2012
Admission routes1
Has abstractyes

Explore more

Same venueUniversity Library - University of Saskatchewan (University of Saskatchewan)Same topicSubtitles and Audiovisual MediaFrench-language works237,207