chenxwh / diffsynth-exvideo

Fast sdxl with higher quality

CogVLM2: Visual Language Models for Image and Video Understanding

Depth estimation with faster inference speed, fewer parameters, and higher depth accuracy.

Updated to OpenVoice v2: Versatile Instant Voice Cloning

Audio-based Lip Synchronization for Talking Head Video

Fast and High-Quality Text-to-video Generation

OmniGen: Unified Image Generation

Scalable Streaming Speech Synthesis with Large Language Models

DiT-based video generation model for generating high-quality videos in real-time

Convert LLM's coding to image generation

Sharp Monocular Metric Depth in Less Than a Second

Minimal and Universal Control for Diffusion Transformer - demo for Subject-driven generation

Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

CogVLM2: Visual Language Models for Image and Video Understanding

Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

A Powerful Multi-Resolution Diffusion Transformer with Fine-Grained Chinese Understanding

Depth Any Video with Scalable Synthetic Data

High-Fidelity Single-Step Diffusion through Dynamic Adversarial Training

Generating Consistent Long Depth Sequences for Open-world Videos

One Diffusion to Generate Them All

Diffusion-based Visual Foundation Model for High-quality Dense Prediction

Efficient Visual Generation with Hybrid Autoregressive Transformer

Minimal and Universal Control for Diffusion Transformer - demo for Spatially aligned control

Image-to-Video Diffusion Models with An Expert Transformer

Finer and Faster Text-to-Image Generation via Relay Diffusion

Text-to-Video Diffusion Models with An Expert Transformer

Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

Autoregressive Video Generation without Vector Quantization

Emu3-Gen for image generation

Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution

Emu3-Chat for vision-language understanding

Autoregressive Image Generation without Vector Quantization

Let Vision Language Models Reason Step-by-Step

Enhancing Video Model Post-Training through Data, Reward, and Conditional Guidance Design

Text-to-Video Diffusion Models with An Expert Transformer