--- name: percom-camera-ready description: Use when preparing an accepted IEEE PerCom paper for its IEEE Xplore camera-ready, covering de-anonymization, the IEEEtran compsocconf template and page budget, IEEE Xplore metadata (PDF eXpress, eCopyright, ORCID, index terms), integrating reviewer-required changes without scope creep, permanentizing the dataset links, and IEEE production checks. --- # PerCom Camera Ready Use this after acceptance. PerCom papers publish in **IEEE Xplore**, so the camera-ready is an **IEEE production step** with its own machinery — PDF eXpress compliance and the IEEE electronic copyright (eCopyright) form — not just a formatting pass. Reopen the current camera-ready instructions, the decision email, and the registration/presentation pages before advising. ## Camera-ready audit - **De-anonymize completely:** restore the author block, affiliations, ORCIDs, acknowledgements, funding, and the real testbed/system/dataset names that double-blind review forced you to hide. - **Apply the final format:** IEEEtran `\documentclass[10pt, conference, compsocconf]{IEEEtran}`, within the current page budget (9+1) plus any camera-ready allowance the acceptance letter states. - **Complete IEEE Xplore metadata:** title/abstract, author order and ORCIDs, and **IEEE index terms / keywords** (and ACM CCS concepts if requested). Journal-grade metadata errors are harder to fix after publication than they look. - **Pass PDF eXpress:** validate the final PDF through IEEE PDF eXpress for Xplore-compatible fonts and embedding; a PDF that fails eXpress will not be accepted for the proceedings. - **Complete the IEEE eCopyright form:** the correct copyright/rights statement must appear on the first page as the form dictates. - **Integrate reviewer-required changes** faithfully — without strengthening claims beyond what was evaluated in the (single-round) review. - **Permanentize open data:** replace every anonymized dataset/code link with a public, licensed, DOI-issuing archive, and update the availability statement to point at it. - **Confirm registration and in-person presentation:** at least one author registers and presents; remote/pre-recorded presentation is not permitted except in stated special circumstances. ## De-anonymization sweep | Anonymized at submission | Restore at camera-ready | Watch for | |---|---|---| | Author block, affiliations, ORCID | Full, correctly ordered | Wrong author order breaks the DOI citation | | Testbed / system / dataset name | Real name throughout text, figures, dataset | A leftover placeholder in a caption or figure | | Acknowledgements, funding | Restored | Grant numbers required by funders | | Self-citations (third person) | Natural first-person where it aids clarity | Over-correcting and double-citing | | Dataset/code link | Public archive DOI (IEEE DataPort / Zenodo) | The old anonymized URL surviving in a footnote | | Sensor-rig / deployment photos | Real photos if desired | A face or building that needs consent to publish | ## IEEE Xplore production checks ```text [Template] IEEEtran compsocconf, current revision; no manual margin/font edits; 10pt two-column [Metadata] title, abstract, authors+ORCID, index terms/keywords entered and matching the PDF [PDF eXpress] final PDF validated and Xplore-compatible (fonts embedded) [eCopyright] IEEE electronic copyright form completed; correct rights line on the first page [References] complete, consistent; DOIs where available; within the 1 reference page [Figures] vector where possible; readable in print greyscale; captions self-contained [Links] every dataset/code link resolves from a logged-out browser ``` ## Worked example: integrating a rebuttal-driven clarification The reviewers asked you to make the cross-subject nature of the headline result explicit. Camera- ready move: reorder §4.1 so the leave-one-subject-out F1 leads and the pooled number is clearly secondary, restore the real system name in that table, point the availability statement at the now-public IEEE DataPort dataset, and add the ORCID and index terms — **without** expanding the claim beyond what the single review round evaluated. ## Dataset-release handoff The camera-ready and any dataset/artifact deposit are related but separate tasks. Do not let the paper's availability statement promise more than the public dataset delivers: align the statement with what you actually deposit (a de-identified dataset under an open license with a DOI), and note honestly anything (raw video, re-identifiable traces) that stays restricted and why (`percom-reproducibility`, `percom-artifact-evaluation`). ## Hedged logistics - Page allowances, metadata fields, PDF eXpress and eCopyright mechanics, and exact camera-ready dates change each cycle; confirm against the decision email and current instructions rather than a prior year. Whether a reproducibility-badge track runs is **待核实** per edition. ## Output format ```text [Camera-ready status] ready / needs fixes / blocked [De-anonymization] author block / system name / acks / links restored? yes/no [IEEE metadata] ORCID / index terms / eCopyright / PDF eXpress passed? yes/no [Reviewer-change map] final edit, no scope creep> [Open data] anonymized links replaced by DOI archive? yes/no [Remaining owner] task> ```