Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Daten und Forschung · SKILL.md

vis-camera-ready

vis-camera-ready

Use when preparing an accepted IEEE VIS paper for its IEEE TVCG camera-ready, covering de-anonymization, the VGTC/TVCG template and journal metadata (DOI, ORCID, IEEE keywords/index terms), integrating the second-round required changes without scope creep, permanentizing open-materials links, the Open Practices disclosure form, and the Graphics Replicabilit…

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: vis-camera-ready description: Use when preparing an accepted IEEE VIS paper for its IEEE TVCG camera-ready, covering de-anonymization, the VGTC/TVCG template and journal metadata (DOI, ORCID, IEEE keywords/index terms), integrating the second-round required changes without scope creep, permanentizing open-materials links, the Open Practices disclosure form, and the Graphics Replicability Stamp handoff.

VIS Camera Ready

Use this after acceptance. IEEE VIS full papers are published as IEEE TVCG journal articles, so the camera-ready is a journal production step, not just a formatting pass. Reopen the current TVCG author instructions, the "Final Information for Paper Authors" page, and the Open Practices form before advising.

Camera-ready audit

  • De-anonymize completely (if you submitted double-blind): restore the author block, affiliations, ORCIDs, acknowledgements, funding, and the real system, tool, and repository names the anonymization hid.
  • Apply the final format: the VGTC / TVCG-conference template within the page budget (9 pages of content + up to 2 reference/acknowledgement pages), with the camera-ready allowances the acceptance letter states.
  • Complete TVCG journal metadata: title/abstract, author order, ORCIDs, IEEE keywords / index terms, and the DOI-bearing citation once assigned. Journal metadata errors are harder to fix after publication than conference ones.
  • Integrate the second-round required changes faithfully — without strengthening claims beyond what the reviewers evaluated.
  • Permanentize open materials: replace every anonymized code/data/video link with a public, licensed, archival location (OSF, Zenodo, a tagged release), and make the paper point at it.
  • Complete the Open Practices form (preprint, preregistration, open data/code, supplemental disclosures) that accepted authors fill in at camera-ready.
  • Confirm presentation obligations; at least one author presents at the conference.

De-anonymization sweep

Anonymized at submissionRestore at camera-readyWatch for
Author block, affiliations, ORCIDFull, correctly orderedWrong author order breaks the DOI citation
System / technique nameReal name throughout text, figures, video, repoA leftover placeholder name in a caption or the video
Acknowledgements, fundingRestoredGrant numbers required by funders
Self-citations (third person)Natural first-person where it aids clarityOver-correcting and double-citing
Supplemental videoRe-render with real title/affiliation slateThe anonymized voice-over or blurred logo surviving
Code/data linkPublic archival DOI (OSF/Zenodo)The old anonymized URL surviving in a footnote

TVCG production checks

[Template]   VGTC / TVCG-conference template, current revision; no manual margin/font edits
[Metadata]   title, abstract, authors+ORCID, IEEE keywords/index terms entered and matching PDF
[References] complete, consistent; DOIs where available; within the 2-page reference allowance
[Figures]    vector where possible; readable in print and grayscale; CVD-safe palettes; captions self-contained
[Rights]     IEEE copyright/eCF form completed; correct rights statement on the first page
[Links]      every open-materials and video link resolves from a logged-out browser
[Video]      supplemental video re-rendered, de-anonymized, correct codec/length

Worked example: integrating a second-round required change

The reviewers required adding the perceptual pilot and rewriting the encoding rationale. Camera-ready move: place the pilot's result with its effect size and CI in the evaluation section, revise §4 so each visual channel is tied to a task, restore the real system name in the figures and the video, and point the open-materials statement at the now-public OSF archive — without expanding the claim the reviewers accepted.

Replicability-Stamp and Open Practices handoff

The camera-ready and the Graphics Replicability Stamp review are separate tracks. Do not wait for the stamp result to finalize the paper, but do make the paper's open-materials statement consistent with what you are submitting to GRSI, and complete the Open Practices disclosures honestly. The stamp is earned by an independent volunteer reproducing your code/data, so the public archive must be the exact, runnable version the paper describes.

Hedged logistics

  • Page allowances, metadata fields, the IEEE rights-form mechanics, and exact camera-ready dates change each cycle; confirm against the decision email and current TVCG instructions rather than a prior year.

Output format

[Camera-ready status] ready / needs fixes / blocked
[De-anonymization] author block / system name / acks / links / video restored? yes/no
[TVCG metadata] ORCID / IEEE keywords / rights form complete? yes/no
[Required-change map] <second-round change -> final edit, no scope creep>
[Open practices] anonymized links replaced by archival DOI? disclosure form done? yes/no
[Remaining owner] <person -> task>
Gespeicherte Skills