About the conference
CVPR is computer vision’s flagship and one of the largest — and by some citation metrics most impactful — publication venues in all of science. Everything visual lives here: recognition, 3D, generative models, video, embodied vision, and the datasets that drive them.
Papers are typically due in November for the June conference, with rebuttals over the winter and decisions in late February. Its odd-year sibling ICCV and even-year European counterpart ECCV give the vision community a second top-tier deadline every year.