Rss Camera Ready

RSS Camera Ready

Turn the anonymous accepted version into the archival one. The 2026 camera-ready deadline was never publicly surfaced (待核实 — take it from the acceptance email); everything else below is verified venue mechanics as of 2026-07-08.

What changes at this stage, specifically at RSS

The links row is the venue-specific one: RSS 2026 explicitly deferred all external linking to the camera-ready, so this is the moment the project page, code repository, and hosted video enter the paper. Test every link logged-out before submitting the final PDF.

Edits that are — and are not — in scope

• In scope: corrections promised in the rebuttal, scope-sentence adjustments the AC asked for, fixed typos, restored acknowledgements, added links.
• Out of scope: new experiments upgrading the claims, retitled contributions, or quietly strengthened numbers. The accepted paper is what the AC meeting endorsed; divergence beyond clarification is an integrity problem, not a formatting one.
• Keep section, table, and figure numbering stable where possible so the review record still maps onto the final text.

Publication mechanics

• RSS proceedings appear at roboticsproceedings.org — free to read, no author fee, no paywall — with DOIs of the form 10.15607/RSS.<year>.<vol>.<paper>. Verify your final metadata (title capitalization, author order, affiliations) exactly as it should be archived; the volume page is what dblp and citation indices ingest.
• There is no APC anywhere in this pipeline; the cost model is conference registration. Whether at least one author must register/present in person was not publicly verified for 2026 (待核实 against the acceptance email).

The single-track presentation

Acceptance at RSS means presenting to the entire conference, not a parallel-session slice — the preparation bar is accordingly higher:

1. Build the talk for the distant-subfield attendee: one slide establishing why your problem matters outside your niche, before any method detail.
2. Show real footage early; a single uncut clip buys more belief than an architecture diagram.
3. Rehearse the failure question — "when does it break?" is the most common single-track question, and your failure table is the prepared answer.
4. Poster/spotlight mechanics vary per edition (待核实 in author instructions); prepare the poster to carry the trial evidence the talk compresses away.

Archive-metadata proof pass

The roboticsproceedings.org volume page is the record dblp, Google Scholar, and citation managers will ingest — proof it like a legal document:

• Author name forms consistent with each author's existing publication record (initials vs full names decide whether citation graphs merge or fork).
• Title capitalization and any math/Unicode rendered as intended.
• Affiliations current at publication time, not at submission time.
• Funding acknowledgements exactly as grant terms require.
• The 10.15607/RSS.* DOI, once minted, is what belongs in your CV and repo citation file — point the repository's citation entry at the proceedings page, not at the arXiv mirror.

Between acceptance and Sydney: an ordered runway

Eleven weeks separated the 2026 decision (April 27) from conference week (July 13-17). A workable ordering:

Final-pass checklist

[ ] Acceptance-email requirements extracted (deadline, format, forms) — sole source
[ ] Author block, funding, acknowledgements restored
[ ] Platform attribution un-anonymized; self-citations re-personalized
[ ] Links added and tested logged-out: repo / project page / video
[ ] Claims diff vs accepted version = clarifications only
[ ] Metadata proofread as-it-will-be-archived
[ ] Public artifact release live before the conference talk drives traffic to it

Hedged logistics (all 待核实 per cycle)

• Camera-ready deadline, template version, and any form uploads: sole source is the acceptance email plus the author-instructions page it links.
• Registration classes, pricing, and whether a paper requires a specific class or in-person presenter: verify on the current registration page.
• Copyright/consent mechanics for the proceedings: RSS's open-access model differs from IEEE-style transfer flows — do not assume either shape; read the form sent.
• Talk length, spotlight-vs-poster split, and session assignment: published with the program, not the acceptance email; recheck close to conference week.

Output format

[Camera-ready status] blocked / in progress / submitted
[Sole-source items] <from acceptance email: deadline, forms>
[Diff audit] clarifications only / out-of-scope edits found
[Link check] <url -> ok/broken>
[Presentation prep] whole-audience slide / footage / failure answer