La interfaz está localizada; las descripciones y documentos de las Skills se traducen automáticamente con IA a su idioma. El código, los comandos, rutas y términos técnicos se mantienen en el original, y prevalece el documento del autor.
Explorar Skills
Datos e investigación · SKILL.md

sensys-camera-ready

sensys-camera-ready

Use when preparing an accepted SenSys camera-ready — de-anonymizing the double-column ACM final, completing ACM rights and metadata, restoring acknowledgments, landing the awarded ACM artifact badges onto the paper, releasing traces and firmware within their constraints, and planning the in-person talk and demo for the merged audience.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentación de la Skill

GitHub

name: sensys-camera-ready description: Use when preparing an accepted SenSys camera-ready — de-anonymizing the double-column ACM final, completing ACM rights and metadata, restoring acknowledgments, landing the awarded ACM artifact badges onto the paper, releasing traces and firmware within their constraints, and planning the in-person talk and demo for the merged audience.

SenSys Camera-Ready

Acceptance flips the paper from anonymous-under-review to a permanent ACM DL record. The camera-ready is a checklist of irreversible steps: de-anonymize completely, get the ACM rights and metadata exactly right, land any artifact badges onto the printed paper, and prepare to present to a room that now spans the merged SenSys/IPSN/IoTDI communities. A metadata or rights slip here is corrected only by a painful post-publication erratum.

De-anonymize completely

Everything double-blind suppressed now goes back — and nothing may be missed:

[ ] Author names, affiliations, and emails restored on the paper.
[ ] Acknowledgments and funding/grant numbers added back.
[ ] Self-citations returned to first person where natural.
[ ] Artifact links de-blinded to their permanent public homes (with DOIs).
[ ] PDF metadata now correctly names authors (was scrubbed for review).

Run the inverse of the submission blindness sweep: the risk now is an un-restored acknowledgment or a still-anonymized repo link, not a leak.

ACM rights and metadata

The ACM e-rights process gates the final. Complete it precisely:

ItemActionFailure cost
e-rights formChoose the license/transfer the process offers; get the exact rights text + DOIWrong rights block on the PDF
Rights stampPaste ACM's rights text and conference string verbatim into the templateReprint / erratum
Author metadataNames, affiliations, ORCIDs match the form and the PDFACM DL record errors
Title/abstractMatch the accepted paper exactlyIndexing mismatch
Template versionUse the ACM double-column version the CFP names (待核实 per cycle)Production rejection

Extra camera-ready pages are sometimes granted for reviewer-requested additions; confirm the exact allowance for the cycle before spending them.

Coordinate the artifact badges onto the paper

If you pursued badges (sensys-artifact-evaluation), the camera-ready is where awarded badges are attached to the printed paper and the ACM DL metadata. Confirm:

[ ] Each awarded badge (Available / Functional / Reproduced) appears on the final PDF.
[ ] The permanent artifact DOI in the paper matches the archived deposit.
[ ] Released traces/firmware honor their consent/NDA/location constraints (sensys-reproducibility).
[ ] The public repo README maps claims → the script/trace that reproduces each figure.

Do not release sensor data, firmware, or deployment locations you have not cleared — the public release is permanent.

Plan the talk and demo

SenSys is in-person; presenting to the merged audience means a room with sensor-networks, embedded, IoT, and on-device-AI expertise all present:

  • Lead with the physical problem, not the architecture — the same first-page arc as the paper (sensys-writing-style), compressed to a few minutes.
  • Show the measurement, not just the claim: a power trace or a latency CDF lands harder than a bullet saying "efficient."
  • A live or recorded demo of the node/deployment is unusually persuasive at SenSys; if you demo, rehearse the failure modes (a dead battery on stage is avoidable).
  • Anticipate the merged audience's questions: an IPSN-rooted reviewer asks about the protocol; an IoTDI-rooted one about deployment; an embedded-AI one about footprint. Have all three ready.

Final camera-ready sequence

1. De-anonymize fully; inverse blindness sweep passes.
2. ACM e-rights complete; rights text + DOI stamped verbatim in the template.
3. Metadata (authors, ORCIDs, title, abstract) matches the form and PDF.
4. Awarded badges + artifact DOI on the final PDF; public release cleared.
5. Compile in the correct template version; re-read the produced PDF.
6. Talk + demo built from the measured-behavior arc; merged-audience Q&A prepared.

Output format

[Deanon]   de-anonymization complete? un-restored items listed
[Rights]   ACM e-rights done; rights text + DOI stamped — pass/gap
[Metadata] authors/ORCIDs/title/abstract consistent across form and PDF
[Badges]   awarded badges + artifact DOI on the final; release cleared
[Talk]     talk/demo built from the measured-behavior arc; Q&A ready
[Open]     the one irreversible item still unconfirmed before the deadline
Skills guardadas