Few-step distillation for autoregressive (AR) video generation commonly first initializes an AR student and then refines it with distribution matching distillation (DMD). This refinement process faces two challenges. First, although initialization already yields a reasonably capable student, DMD obtains the refinement...
Zi-Xuan Duan, Ya-Bo Chen, Xun-Zhi Xiang et al.· 0 citations
Continuous visual world simulation requires coupling evolving scene geometry with temporally persistent appearance. We present UniWorld, a unified framework that integrates depth trajectories and reference appearance within a shared video generation model. The central design maintains visual and geometric histories on...
Xin Zhang, Ya-Bo Chen, Shuai-Wei Wang et al.· 0 citations
Visual autoencoders must balance reconstruction fidelity, compactness, and semantic structure. Existing semantic tokenizers typically inherit semantics from pretrained recognition encoders, whose invariances can discard the appearance detail required for inversion. We show that this inheritance is unnecessary: semantic...
Yu-Zhou Ji, Jie-Peng Wang, Tian-Yu Jin et al.· 0 citations
This work proposes ReCam, which groups local spatiotemporal blocks of historical latents into vectors and projects them directly into the high-dimensional hidden space of a Diffusion Transformer, which learns to use the resulting memory tokens through native self-attention, reducing token count while retaining block-le...
Yu-Kuo Ma, Cong Liu, Junke Wang et al.· 0 citations
Few-step distillation improves the efficiency of autoregressive (AR) video generation, but often causes diversity collapse: under the same prompt, different noise samples tend to produce highly similar videos with weakened motion dynamics. We analyze this degradation in Distribution Matching Distillation (DMD)-distille...
Zi-Xuan Duan, Xun-Zhi Xiang, Yabo Chen et al.· 0 citations
We present DistScene, a framework for single-image compositional 3D scene generation by jointly modeling the environment and individual objects. Unlike existing methods that represent scenes primarily as collections of objects, we model the environment as an explicit scene component to provide geometric context for obj...
Kun-Ming Luo, Hong-Yu Yan, Ken Deng et al.· 0 citations
Unified multimodal models integrate visual understanding and generation within a single network, yet the two capabilities are commonly optimized as separate tasks. We introduce Generative Grounding Feedback(GGF), a self-evolving post-training framework that uses only text prompts and the model's own visual experience....
World Action Models (WAMs) augment robot action generation with future visual supervision. Existing WAMs commonly fuse main and wrist observations into one visual stream and train both with the same future-video objective, despite their different visual dynamics. A stable main camera reveals scene-level task evolution,...
Jie Wu, Yu-Zhi Huang, Jun-Qi Liu et al.· 0 citations
TourPhysics combines deterministic simulation with video generation while assigning separate roles to simulator state, geometric evidence, generator controls, and appearance memory, an online framework initialized from a single image and a declarative physical configuration.
Xin Zhang, Yabo Chen, Zi-Xuan Duan et al.· 2 citations
A unified conditional-flow framework that casts generation, semantic editing, and intra-structural retargeting as condition-modulated transport within one text- and skeleton-conditioned rectified-flow model, replacing separate pipelines with a single conditional motion transport model while keeping the same-topology re...
Junli Li, Xin Song, Si-Qi Wang et al.· arXiv.org· 1 citation
Experimental results show that GenTrans supports effective video transmission under ultra-low-bitrate and weak-network conditions, achieving improved transmission efficiency, decoding efficiency, and robustness while preserving perceptual quality.
Xiangyu Chen, Ji-Xiang Luo, Yuan-Kai Fan et al.· arXiv.org· 2 citations
This paper proposes a unified framework named CineWeaver to achieve reference-controllable multi-shot long-video generation without retraining, and is the first unified framework to simultaneously enable long-form, reference-controllable, and multi-shot video generation in a training-free fashion.
Yuyang Huang, Yabo Chen, Wen-Rui Dai et al.· arXiv.org· 4 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.