Signals 4 · free daily AI digest

DynaTokens: Teaching Dynamics to Camera-Controlled Video Models at Test Time

Paper recorded by Signals 4 on 2026-09-28 in cs.CV. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-28 on arXiv · recorded by Signals 4 on 2026-09-29

Category: cs.CV · 计算机视觉 · first seen 2026-09-29

Abstract

Video generation must account for two sources of motion, one induced by the observer's camera path and the other caused by scene dynamics. An ideal camera-controlled video model should account for both motions: let users move the camera while evolving the scene dynamics. While current models handle camera-induced motion well in static settings, they struggle for dynamic scenes: objects are static,

Read on arXiv →

#10 most recent of 342 cs.CV papers we have recorded · ↑ newer: Mind the RefGAP: Correcting Reference Attention in Diffusion-Based Vis · ↓ older: FlowTool: Controlling Tool Parameter in Image Retouching via Flow Matc
Cite this page: DynaTokens: Teaching Dynamics to Camera-Controlled Video Models at Test Time: the #10 most recent of 342 cs.CV papers we have recorded (as of 2026-09-28). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/dynatokens-teaching-dynamics-to-camera-controlled-video-models-at-test-time.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable: papers.json
Related: More cs.CV papers · arXiv signals · All papers · Today in AI
Get 4 AI signals a day by email — free.
Subscribe free → See all plans →
Get 4 AI signals a day by email — free
All models · All repos · By company · Daily editions