Awesome-RLHF-Video-Diffusion

Table of Contents

• Base Model with RLHF
• DPO
• GRPO
• Reward Guidance

Base Model with RLHF

Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model

📄 Paper

Seedance 1.0: Exploring the Boundaries of Video Generation Models

📄 Paper

Skyreels-v2: Infinite-length film generative model

📄 Paper | 💻 Code

Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Model

📄 Paper | 🌐 Project Page | 💻 Code

DPO

Improving Video Generation with Human Feedback

📄 Paper | 🌐 Project Page | 💻 Code

VideoDPO: Omni-Preference Alignment for Video Diffusion Generation

📄 Paper | 🌐 Project Page | 💻 Code

DenseDPO: Fine-Grained Temporal Preference Optimization for Video Diffusion Models

📄 Paper

AlignHuman: Improving Motion and Fidelity via Timestep-Segment Preference Optimization for Audio-Driven Human Animation

📄 Paper | 🌐 Project Page

Hallo4: High-Fidelity Dynamic Portrait Animation via Direct Preference Optimization and Temporal Motion Modulation

📄 Paper | 💻 Code

LiFT: Leveraging Human Feedback for Text-to-Video Model Alignment

📄 Paper | 🌐 Project Page | 💻 Code

GRPO

MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE

📄 Paper | 🌐 Project Page | 💻 Code

DanceGRPO: Unleashing GRPO on Visual Generation

📄 Paper | 🌐 Project Page | 💻 Code

Flow-GRPO: Training Flow Matching Models via Online RL

📄 Paper | 🌐 Project Page | 💻 Code

Reward Guidance

GigaVideo-1: Advancing Video Generation via Automatic Feedback with 4 GPU-Hours Fine-Tuning

📄 Paper | 🌐 Project Page | 💻 Code