Tag
This paper introduces IllusionReasoning, a benchmark using real-world visual illusions to jointly evaluate the perception and reasoning capabilities of Large Vision Language Models (LVLMs), finding that current models' reasoning abilities are not as advanced as claimed.
JanusMesh is a fast, training-free framework that generates text-driven 3D visual illusions—a single mesh revealing different semantics from different viewing angles—by decoupling generation into cross-space dual-branch denoising and view-conditioned texture synthesis, achieving high realism in just 3-5 minutes.