SketchVerify: Physics-Savvy Planning for More Realistic AI Videos

Kari Jaaskelainen

24 Nov 2025 — 1 min read

Ever asked AI to make a video and got floating cups or jittery motion? SketchVerify is a training-free planning trick that makes AI videos more physically believable—before you spend compute on full synthesis.

How it works:

Given a prompt and reference image, it proposes many candidate object trajectories (plans).
Each plan is rendered as a lightweight “video sketch” by compositing moving objects over a static background—no heavy diffusion needed.
A vision-language verifier scores these sketches for instruction match and physical plausibility, picks the best, and iteratively refines until good enough.
The chosen plan then guides the final trajectory-conditioned generator—just once.

Results: On WorldModelBench and PhyWorldBench, SketchVerify improves motion quality, realism, and long-term consistency while being significantly more efficient. More candidate trajectories consistently yield better outcomes.

Paper: https://arxiv.org/abs/2511.17450v1

Authors: Yidong Huang et al.

Paper: https://arxiv.org/abs/2511.17450v1

Register: https://www.AiFeta.com

#AI #VideoGeneration #GenerativeAI #ComputerVision #Physics #Planning #Diffusion #Research

Meet GRACE: a moral governor for safer, more transparent AI

AI agents are getting powerful—so how do we make sure they do the right thing, not just the effective thing? Meet GRACE, a reason-based moral governor that keeps AI behavior aligned with human norms by separating moral reasoning from goal-driven decision-making. * Moral Module: uses deontic logic and explicit reasons

Why some fine-tuned LLMs miss phishing—and how to fix it

Not all fine-tuned LLMs spot phishing equally. A new study tests Llama 3.1 8B, Gemma 2 9B, and Mistral on high-stakes phishing detection—and uses SHAP and mechanistic interpretability to reveal why models do (or don’t) generalize. * Architecture × data diversity matters: Gemma 2 9B hits state-of-the-art performance (F1

AI that explains itself—by following the science

AI that explains itself—by following the science Black-box AI can be powerful, but it often can’t tell us why it made a decision. Concept Bottleneck Models (CBMs) try to fix that by predicting human-understandable concepts first, then the final answer. The catch: standard CBMs ignore domain-specific cause-and-effect and

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

TL;DR: Finding reliable answers in genomics is hard. GenomAgent turns one big AI into a coordinating team of specialists—and beats the current leader by 12% on a key benchmark. Genomic facts live across many databases. Standard chatbots struggle because they can’t flexibly query those sources. GeneGPT added

Read more

Meet GRACE: a moral governor for safer, more transparent AI

Why some fine-tuned LLMs miss phishing—and how to fix it

AI that explains itself—by following the science

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A