DreamFree Repurposes Video Diffusion for World Action Models
DreamFree enables real-time robot control at 11Hz without generating future videos.
Stories per day · 30 days
2 RSS
Relatedinference3world-models34d-reconstruction2attention2generative-models2hunyuanvideo2robotics2world-action-models23d-gaussians1autoregressive1camera-control1distillation1dreamx1embodiment-transfer1fred1glance-forcing1imagination-free1latent-action1
22 stories, newest first · one card per story
Stories tagged #video-diffusion
DreamX repurposes video diffusion models as universal latent action learners
DreamX extracts 300x compressed latent actions from pretrained video diffusion models, enabling any-to-any video transfer and zero-shot action transfer across embodiments.
DreamX Video Diffusion Model as Universal Latent Action Learner
DreamX extracts compact latent actions from video diffusion models for transfer across embodiments and environments.
AB-Attention approximates dense video attention at high sparsity
On Wan2.1 and HunyuanVideo the authors report skipping 89–92% of attention with up to 2.23× speedup.
L3G-World generates consistent worlds in real time
Latent 3D Gaussian memory grows with scene extent, not stream length, and the authors report 9 FPS generation.
FRED improves future video representations for world action models
On tasks shown only in human videos, FRED reaches 50.0% success versus 19.4% for the best rival.
ALLOT allocates block-sparse attention for video diffusion transformers
Training-free ALLOT gains 11.8 dB PSNR on HunyuanVideo with XAttention at equal or shorter runtime.
RealMe builds 2K 4D-consistent volumetric video from sparse captures
A high-resolution video diffusion model refines limited multi-view captures and beats prior methods on DNA-Rendering and RealVV.
Holo4D steers video diffusion with reconstructed 4D trajectories
Projected source-view trajectories condition a pretrained video model so unseen regions are filled without retraining the generator.
Velox: Spatial Memory Makes World Models Efficient
Velox speeds long video world-model rollouts 2.43x at chunk 48 while staying near dense quality.
Glance Forcing: Making Bidirectional Diffusion Autoregressive at a Glance
Glance Forcing distills causal video diffusion with one PF-ODE trajectory and dual slow-fast experts.
ForcingBooster: Adaptive Patch-wise Denoising for Fast Autoregressive Video Diffusion
ForcingBooster cuts autoregressive video diffusion steps about 50% by stopping easy patches early.
DeCoPrune: Efficient KV-Cache Pruning for Autoregressive Video Diffusion via Denoising Consistency
DeCoPrune cuts over 85% of video-diffusion KV tokens while preserving long-range recall.
Breaking the Uniformity Trap: Scaling Video Diffusion Model via SplitMoE
SplitMoE splits video-diffusion experts into semantic and generic roles, improving quality under a fixed parameter budget.
Breaking the Uniformity Trap: Scaling Video Diffusion Model via SplitMoE
SplitMoE divides video-diffusion experts into semantic and generic roles, avoiding uniform routing and improving generation quality.
Strike a Chord! Modal Kinetic Typography
Researchers animate vector glyphs from vibration modes, with a frozen video diffusion model controlling amplitudes and phases.
LongLive-Plug: Once-for-All Distillation for Video Generation
LongLive-Plug distills reusable LoRAs so 54 video models gain sampling and guidance capabilities.
PDMD: Projected Distribution Matching Distillation for Video Diffusion Models
PDMD stabilizes video diffusion distillation by filtering critic errors, beating DMD at 4 NFE.
Sol-H3: Recursive Self-Improvement for MiniMax-H3 Inference Acceleration on Sol-Engine across Cloud and Edge
Sol-H3 accelerates 33B MiniMax-H3 video diffusion up to 30×, beating real time on an 8×GB200 node.
In-Flight KV Cache with Clean Anchors for Faster Autoregressive Video Diffusion
FlashForward reuses in-flight KV caches and clean anchors to speed long autoregressive video diffusion.
Streaming Video Editing with Easy Adaptation
SVEET adapts a bidirectional video diffusion model for real-time streaming edits at 15 FPS on one H100.
Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanisms
Researchers trace unphysical video-diffusion motion to RoPE attention decay and propose a frequency-scaling fix.