TaoMate-R1: Temporal Role Specialization for Streaming Multi-Reference Video Generation
cs.CV
Submitted: 2026-11-10
Updated: 2026-11-10
Code: https://github.com/TaoLiveAIGC/TaoMate-H3
Terminology
Sources
- Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion
- Self Gradient Forcing: Native Long Video Extrapolation
- SGF+: Decoupling Gradient Flows for Autoregressive Video Generation
- DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control
- FlowAct-R2: Beyond Talking Avatar via Streaming Multimodal References and Proactive Agent Planning
- ReferStream: Interactive Subject Injection for Streaming Video Generation
- Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion
- SkyReels-V2: Infinite-length Film Generative Model
- LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generation
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- Rolling Forcing: Autoregressive Long Video Diffusion in Real Time
- eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers
- LoRA: Low-Rank Adaptation of Large Language Models
- Improved Distribution Matching Distillation for Fast Image Synthesis
- MultiRef-Compass: Towards Comprehensive Evaluation of Multi-Reference-to-Audio-Video Generation
- DINOv2: Learning Robust Visual Features without Supervision
- SkyReels-V3 Technique Report
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs