Today AAAI2024 sent out notifications of Phase-I rejection. Some of my papers with collaborators received a rejection. A paper with a student of mine did not. She was happy.
But it does not mean anything, Phase-I rejected or not. I said.
Three years ago, about this time of the year, I had a paper submitted to AAAI, which got Phase-I rejected. We did not make any modification and submitted it directly to CVPR, it got accepted. It is this paper:
In this paper, we propose “adversarial model perturbation”, or AMP, which is identical to “sharpness aware minimization’’, or SAM. SAM was published by Google in ICLR2021, a few months earlier than AMP. If AAAI did not Phase-I reject our paper, our paper would have been published first. But now SAM is well known, and very few know about AMP.
This is the cost of getting Phase-I rejected. But I do not blame such a review system: when a conference gets too big, noisy reviews are inevitable.
Another reason AMP is less well known than SAM is that I do not tweet (or shall I say, I do not X?). It is always the loud who get attention. That is another law of mankind.