How to address novelty concerns in top ai conference? [D]

Hi, I’m a researcher working in computer vision. Over the past few years, I’ve submitted several papers to top-tier conferences such as NeurIPS, ICLR, and CVPR, and one concern that seems to come up repeatedly is 'novelty'. Given that thousands of papers are published every year at top conferences alone, not to mention the tens of thousands published across other conferences and journals, I sometimes wonder how much genuinely new novelty is realistically left to explore. In such a crowded research landscape, how do you usually address novelty concerns from reviewers? More specifically, I would really appreciate any advice on how to frame a contribution so that its novelty is clear, how to distinguish meaningful incremental progress from work that may be considered insufficiently novel, and what reviewers generally look for when judging novelty. Any tips or experiences would be greatly appreciated. Thanks!