Robotics ◆ Robot learning ◆ Machine learning ◆ Vision ◆ Checked against the official calls
Every venue I might send a paper to, on one clock: when the paper is due, when the decision comes back, and when and where the conference meets. The planner below chains one venue's decision into the next deadline, so a rejected manuscript always has somewhere to go.
Today is —. Countdowns are in whole days to the deadline date; most deadlines close at 23:59 Anywhere-on-Earth, which is 07:59 the next morning in New York.
One row per conference edition, all on the same time axis. A red diamond is the paper deadline, the dashed red stretch is the paper under review, the gold dot is the decision, and the blue block is the meeting itself. The gap between a gold dot on one row and a red diamond further right on another is a resubmission window. Click a row for the full schedule and its sources.
The rule is simple: a paper can only be under review at one venue at a time, so the useful question is which deadline comes after the decision you are waiting for. If RSS decides today and ICRA closes next week, the same manuscript can go to RSS first and, if it is turned down, straight on to ICRA. The planner finds those windows and lines them up into a ladder.
Decisions that land before each deadline
For every upcoming deadline in the next nine months: the venues whose decision you will already have. Green means there is at least the buffer to revise; amber means the decision lands inside the buffer; red means it lands after the deadline.
Deadlines that collide
Two deadlines within ten days of each other cannot take the same manuscript. Either send two different papers or choose one.
Rules of the road
No concurrent submission. ICRA, IROS, RSS, CoRL, ICLR, NeurIPS, ICML and the CVF conferences all forbid submitting a paper that is under review elsewhere, and the ban covers substantially similar papers. Withdraw first, then resubmit.
Abstract registration is a hard gate. ICLR, NeurIPS, ICML and CVPR require an abstract a few days to two weeks before the paper. Miss it and the ladder skips that rung.
Workshops are usually exempt. Workshop papers and RA-L or T-RO journal submissions follow their own rules; check the venue's dual-submission policy for the exact wording.
arXiv is fine almost everywhere, but double-blind venues ask you not to advertise the preprint during review.
Every venue with its typical cycle, the next edition, and the last few years as a reference. The filter bar above applies here too, so you can narrow to a field, a continent or a deadline season.
Each edition was checked against its official call for papers or important-dates page, and the supporting sentence is quoted in the edition's sources (open any row and expand Sources). Aggregators such as aideadlin.es and ccfddl were used only to cross-check, never as the source of a date; where they disagreed with the official page, the official page won and the disagreement is recorded in the notes.
assets/data/deadlines.json, built from the research files in assets/data/deadlines_src/ by scripts/build_deadlines.py, which checks that every edition's dates are in order (abstract → paper → reviews → rebuttal → decision → camera-ready → conference) before writing the page's data.Dates last verified —. Deadlines move; before you plan a week around one, confirm it on the official site linked from every card.