Enregistré dans:
Détails bibliographiques
Auteurs principaux: Chen, Hanlin, Wei, Jiaxin, Song, Xibin, Wang, Yifu, Wang, Steve, Li, Hongdong, Ji, Pan, Lee, Gim Hee
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2605.30855
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866913178922778624
author Chen, Hanlin
Wei, Jiaxin
Song, Xibin
Wang, Yifu
Wang, Steve
Li, Hongdong
Ji, Pan
Lee, Gim Hee
author_facet Chen, Hanlin
Wei, Jiaxin
Song, Xibin
Wang, Yifu
Wang, Steve
Li, Hongdong
Ji, Pan
Lee, Gim Hee
contents Frame-wise action-controlled image-to-video generation is a promising paradigm for interactive world simulation, where each control signal should elicit an immediate visual response. However, maintaining visual fidelity and 3D consistency over long autoregressive rollouts remains challenging. Existing 3D-aware methods often suffer from catastrophic drift due to two impediments: information loss from \textit{Latent--RGB Cycling}, where generated latents are repeatedly decoded to RGB and re-encoded for future conditioning, and the training--inference gap induced by the \textit{error-free hypothesis}, where clean training memory fails to match prediction-corrupted inference memory. To address these challenges, we present \textbf{Robust Dreamer}, a memory-augmented framework built around how to design 3D memory and how to use it robustly. First, we introduce \textbf{Latent Gaussian Memory}, which anchors diffusion latents inherited from the generation process to Gaussian primitives and recalls them via latent-space Gaussian splatting. This provides dense, geometry-aware, view-aligned conditioning while avoiding accumulated degradation from repeated VAE conversion. Second, we propose \textbf{Deviation Learning with Dynamic Deviation Archive}, which synthesizes rollout-induced latent deviations through a one-step approximation, stores them by autoregressive stage and denoising timestamp, and injects them into historical memory during training. This exposes the generator to realistic corrupted memory states and teaches internal correction before inference. Experiments on ScanNet, DL3DV, and OmniWorldGame demonstrate state-of-the-art long-horizon performance.
format Preprint
id arxiv_https___arxiv_org_abs_2605_30855
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Robust Dreamer: Deviation-Aware Latent Gaussian Memory for Action-Controlled AR Video Generation
Chen, Hanlin
Wei, Jiaxin
Song, Xibin
Wang, Yifu
Wang, Steve
Li, Hongdong
Ji, Pan
Lee, Gim Hee
Computer Vision and Pattern Recognition
Frame-wise action-controlled image-to-video generation is a promising paradigm for interactive world simulation, where each control signal should elicit an immediate visual response. However, maintaining visual fidelity and 3D consistency over long autoregressive rollouts remains challenging. Existing 3D-aware methods often suffer from catastrophic drift due to two impediments: information loss from \textit{Latent--RGB Cycling}, where generated latents are repeatedly decoded to RGB and re-encoded for future conditioning, and the training--inference gap induced by the \textit{error-free hypothesis}, where clean training memory fails to match prediction-corrupted inference memory. To address these challenges, we present \textbf{Robust Dreamer}, a memory-augmented framework built around how to design 3D memory and how to use it robustly. First, we introduce \textbf{Latent Gaussian Memory}, which anchors diffusion latents inherited from the generation process to Gaussian primitives and recalls them via latent-space Gaussian splatting. This provides dense, geometry-aware, view-aligned conditioning while avoiding accumulated degradation from repeated VAE conversion. Second, we propose \textbf{Deviation Learning with Dynamic Deviation Archive}, which synthesizes rollout-induced latent deviations through a one-step approximation, stores them by autoregressive stage and denoising timestamp, and injects them into historical memory during training. This exposes the generator to realistic corrupted memory states and teaches internal correction before inference. Experiments on ScanNet, DL3DV, and OmniWorldGame demonstrate state-of-the-art long-horizon performance.
title Robust Dreamer: Deviation-Aware Latent Gaussian Memory for Action-Controlled AR Video Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.30855