Extraction of title from "section header" pages#102
Open
christianabbet wants to merge 29 commits intodevelopfrom
Open
Extraction of title from "section header" pages#102christianabbet wants to merge 29 commits intodevelopfrom
christianabbet wants to merge 29 commits intodevelopfrom
Conversation
what is a title in the end
…om/swisstopo/swissgeol-assets-dataextraction into feat/issue-98/title-section-header
AgathaSchmidt
requested changes
Mar 9, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Close #98
Description
In this PR we include a title estimation for pages that are labeled as section headers. The ground truth in
data/gt_single_pages.jsonis updated with title entries. To extract the title GT, we used the Pixtral model and a predefined prompt. The Pixtral interface was refactored so that it can be used in the future for other tasks. For title detection we send the image along with the extracted text. Pixtral seems to have poor OCR capability.Moreover, we started some refactoring of the evaluation process. We opened a new issue for future cleanup (#103)
Performance
The title is estimated based on multiple criteria and achieves 29.05% F1-score. For each criterion, the cumulative score is indicated.
API response