TRAC: Trajectory-aware Reuse and Adaptive Correction for Efficient Autoregressive Video Generation
cs.CV
Submitted: 2026-10-02
Updated: 2026-10-15
Comments: Preprint under review
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- Natural Supersymmetry from Extra Dimensions
- SkyReels-V2: Infinite-length Film Generative Model
- ViD-GPT: Introducing GPT-style Autoregressive Generation in Video Diffusion Models
- Ca2-VDM: Efficient Autoregressive Video Diffusion Model with Causal Generation and Cache Sharing
- Efficient Autoregressive Video Diffusion with Dummy Head
- Classifier-Free Diffusion Guidance
- Forcing-KV: Hybrid KV Cache Compression for Efficient Autoregressive Video Diffusion Models
- VideoPoet: A Large Language Model for Zero-Shot Video Generation
- FreqForcing: Autoregressive Long Video Generation via Spectral Self-Anchoring
- Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model
- FreeLong++: Training-Free Long Video Generation via Multi-band SpectralFusion
- FadeMem: Distance-Aware Memory Consolidation for Autoregressive Video Diffusion
- Future Forcing: Future-aware Training-free KV Cache Policy for Autoregressive Video Generation
- Light Forcing: Accelerating Autoregressive Video Diffusion via Sparse Attention
- Flow caching for autoregressive video generation
- HeadCast: Casting Attention Heads for Efficient Autoregressive Video Generation
- MAGI-1: Autoregressive Video Generation at Scale
- Phenaki: Variable Length Video Generation From Open Domain Textual Description
- Sparse VideoGen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity
- Motion-Aware Caching for Efficient Autoregressive Video Generation
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models