Michael Black at ECCV: CV papers are already two years behind frontier models

FinanceYF5 · x · 2026-09-18

Computer vision researcher Michael Black offered a sharp take at ECCV: papers presented at the conference today may already be two years behind frontier models — and in the AI era, two years is enough to render a line of research irrelevant. Starting from his own VIGA paper, he asks what academic computer vision research should still do as general-purpose models keep getting stronger.

Related event: Post-GPT-6 Gloom at ECCV: CV Papers Already Two Years Behind(6 posts)→

Original post →

More from AGI Musings

AGI Musings channel →