TGuide: Faster Autoregressive Video Diffusion via Historical Trajectory Guidance
cs.CV
Submitted: 2026-10-10
Updated: 2026-10-10
Project page: https://zhuochen02.github.io/tguide
Sources
- SkyReels-V2: Infinite-length Film Generative Model
- $\Delta$-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers
- Context Forcing: Consistent Autoregressive Video Generation with Long Context
- Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion Transformers
- Infinite Worlds with Versatile Interactions
- Matrix-game 2.0: An open-source, real-time, and streaming interactive world model
- Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion
- HarmoniCa: Harmonizing Training and Inference for Better Feature Caching in Diffusion Transformer Acceleration
- Autoregressive Adversarial Post-Training for Real-Time Interactive Video Generation
- FreqCa: Accelerating Diffusion Models via Frequency-Aware Caching
- Streaming Autoregressive Video Generation via Diagonal Distillation
- RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO
- Flow caching for autoregressive video generation
- MagCache: Fast Video Generation with Magnitude-Aware Cache
- MAGI-1: Autoregressive Video Generation at Scale
- FORA: Fast-Forward Caching in Diffusion Transformer Acceleration
- WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling
- Matrix-Game 3.0: Real-Time and Streaming Interactive World Model with Long-Horizon Memory
- 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