Jinbo Xing

i1. Code release for "i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models"

live-vlm-webui. Real-time Vision Language Model interaction via webcam - WebRTC-based web interface

world-tracing. Office inference code for World Tracing (object/scene/dynamic). Live demos: https://haoz19.github.io/world-tracing-page/

LLaVA-OneVision-2. Fully Open Framework for Democratized Multimodal Training

diffusion-rm. The official code of "Beyond VLM-Based Rewards: Diffusion-Native Latent Reward Modeling"

DreamVideo-Omni. DreamVideo-Omni: Omni-Motion Controlled Multi-Subject Video Customization with Latent Identity Reinforcement Learning

PiD. PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion

WBench. WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluation

vscode-office. Let VSCode support previewing PDF, Excel, Word and other formats, and add markdown WYSIWYG editor.

Megatron-LM. Ongoing research training transformer models at scale

nba_games. This dataset provides metadata, official statistics, and official play-by-play annotations for full-length NBA game videos available on YouTube. Instead of redistributing video files, we provide YouTube video IDs and URLs so users can download videos independently when their use case and local policies allow it.

DreamX-World. DreamX-World: A General-Purpose Interactive World Model

SwiftI2V. [arXiv 2026] Project page for paper "SwiftI2V: Efficient High-Resolution Image-to-Video Generation via Conditional Segment-wise Generation"

evil-read-arxiv. Claude Code+Obsidian，邪修读论文就是快

tuna-2. Official implementation of Tuna-2: Pixel Embeddings Beat Vision Encoders for Unified Understanding and Generation

MultiWorld. Official Implementation of MultiWorld: Scalable Multi-Agent Multi-View Video World Models

sapiens2. 1K resolution vision transformers pretrained on 1B human images.

nova3r. [ICLR 2026] NOVA3R: Non-pixel-aligned Visual Transformer for Amodal 3D Reconstruction

VTP. [ECCV 2026] Towards Scalable Pre-training of Visual Tokenizers for Generation

VCC. Compile agent conversations!

HyDRA. Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models

seoul-world-model. Seoul World Model: Grounding World Simulation Models in a Real-World Metropolis

Action100M. A Large-scale Video Action Dataset

Kiwi-Edit. A unified and fully open-source framework for instruction-guided and reference-guided video editing using natural language.

BitDance. BitDance & UniWeTok: Open-source autoregressive model with binary visual tokens. A research project for building powerful multimodal autoregressive model.

awesome-notebookLM-prompts. A curated collection of the strongest NotebookLM slide prompts sourced from the real creative underground . Your go-to resource for AI powerpoint :P

ProMoE. [ICLR2026] The official code of "Routing Matters in MoE: Scaling Diffusion Transformers with Explicit Routing Guidance"

DeepSeek-OCR-2. Visual Causal Flow

lingbot-world. Advancing Open-source World Models

self-refine-video. [ICML 2026] Pytorch implementation of Self-Refining Video Sampling

Scale-RAE. Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders

PrismLayers. PrismLayers: Open Data for High-Quality Multi-Layer Transparent Image Generative Models

GLM-Image. GLM-Image: Auto-regressive for Dense-knowledge and High-fidelity Image Generation.

Open-OmniVCus. OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions (NeurIPS 2025)

TurboDiffusion. TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

SVG-T2I. [Arxiv 2025] Official PyTorch Implementation of "SVG-T2I: Scaling up Text-to-Image Latent Diffusion Model Without Variational Autoencoder".

TRELLIS.2. Native and Compact Structured Latents for 3D Generation

Wan-Move. [NeurIPS 2025] Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance

DeepSeek-OCR. Contexts Optical Compression

Ditto. [CVPR'26 Highlight] Ditto: Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset

pico-banana-400k. Python

UniPic. Open-source SOTA multi-image editing model

ISG. (ICLR 2025 Spotlight) Official code repository for Interleaved Scene Graph.

DreamRelation. [ICCV2025] The official code of "DreamRelation: Relation-Centric Video Customization"

OpenING. Official Implementation of OpenING: A Comprehensive Benchmark for Judging Open-ended Interleaved Image-Text Generation

LLM-I. 🚀 LLM-I: Transform LLMs into natural interleaved multimodal creators! ✨ Tool-use framework supporting image search, generation, code execution & editing

Awesome-Unified-Multimodal-Models. Awesome Unified Multimodal Models

Emu3.5. Native Multimodal Models are World Learners

apt. Public release of the code for "Accelerating Vision Transformers with Adaptive Patches"

RISEBench. [NIPS 2025 DB Oral] Official Repository of paper: Envisioning Beyond the Pixels: Benchmarking Reasoning-Informed Visual Editing

Kris_Bench. [NIPS 25'] Evaluation code of paper "KRIS-Bench: Benchmarking Next-Level Intelligent Image Editing Models"

PICABench. PICABench: How Far Are We from Physically Realistic Image Editing?

batrack. [ICCV 2025 Oral] Back on Track: Bundle Adjustment for Dynamic Scene Reconstruction (BA-Track)

RAE. Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

QeRL. [ICLR 2026]QeRL enables RL for 32B LLMs on a single H100 GPU.

Open-DiffusionGS. Baking Gaussian Splatting into Diffusion Denoiser for Fast and Scalable Single-stage Image-to-3D Generation and Reconstruction (ICCV 2025)

ChronoEdit. [ICLR 2026] ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation

SSDD. Official implementation for SSDD Single-Step Diffusion Decoder for Efficient Image Tokenization.

SRPO. Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference

VChain. [ACL 2026 Findings, ICCV 2025 Workshop Outstanding Paper Award] VChain: Chain-of-Visual-Thought for Reasoning in Video Generation

LongLive. Long Video Gen Infrastructure

UniCoT. [ICLR 2026] Uni-CoT: Towards Unified Chain-of-Thought Reasoning Across Text and Vision

DeepResearch. Tongyi Deep Research, the Leading Open-source Deep Research Agent

SpatialVID. [CVPR 2026] SpatialVID: A Large-Scale Video Dataset with Spatial Annotations

LaViDa. Official Implementation of LaViDa: :A Large Diffusion Language Model for Multimodal Understanding

Awesome-Interleaving-Reasoning. Interleaving Reasoning: Next-Generation Reasoning Systems for AGI

Interleaving-Reasoning-Generation. [ICLR 2026] This is an early exploration to introduce Interleaving Reasoning to Text-to-image Generation field and achieve the SoTA benchmark performance. It also significantly improves the quality, fine-grained details and aesthetic aspects of generated images.

PlayerOne. PlayerOne: Egocentric World Simulator

HunyuanImage-2.1. HunyuanImage-2.1: An Efficient Diffusion Model for High-Resolution (2K) Text-to-Image Generation​

dinov3. Reference PyTorch implementation and models for DINOv3

HPSv3. Official implementation of HPSv3: Towards Wide-Spectrum Human Preference Score (ICCV2025)

gpt-oss. gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

VisualSphinx. Python

MotionCanvas. [SIGGRAPH 2025] MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation

Waver. Industry-level video foundation model for unified Text-to-Video (T2V) and Image-to-Video (I2V) generation.

Qwen-Image. Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.

DiffDoctor. [ICCV 2025] DiffDoctor: Diagnosing Image Diffusion Models Before Treating

HunyuanWorld-1.0. Generating Immersive, Explorable, and Interactive 3D Worlds from Words or Pixels with Hunyuan3D World Model

Wan2.2. Wan: Open and Advanced Large-Scale Video Generative Models

TEXGen. [SIGGRAPH Asia 2024, Best Paper Honorable Mention] This is the official implementation of our SIGGRAPH Asia journal artical: TEXGen: a Generative Diffusion Model for Mesh Textures

Long-RL. Long-RL: Scaling RL to Long Sequences (NeurIPS 2025)

True-Story-of-Pangu. 诺亚盘古大模型研发背后的真正的心酸与黑暗的故事。

Reflect-DiT. Reflect-DiT: Inference-Time Scaling for Text-to-Image Diffusion Transformers via In-Context Reflection

metaquery. Official Implementation of Paper Transfer between Modalities with MetaQueries

UniReal. Jupyter Notebook

ml-flextok. FlexTok: Resampling Images into 1D Token Sequences of Flexible Length

Ovis-U1. An unified model that seamlessly integrates multimodal understanding, text-to-image generation, and image editing within a single powerful framework.

qwen2vl-flux. Python

Paper2Poster. [NeurIPS 2025] Open-source Multi-agent Poster Generation from Papers

HDR-GS. "HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting" (NeurIPS 2024)

MiniMax-M1. MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model.

MoviiGen1.1. MoviiGen 1.1: Towards Cinematic-Quality Video Generative Models

MVImgNet2.0. Python

vjepa2. PyTorch code and models for VJEPA2 self-supervised learning from video.

Self-Forcing. Official codebase for "Self Forcing: Bridging Training and Inference in Autoregressive Video Diffusion" (NeurIPS 2025 Spotlight)

RL-GPT. Python

MindOmni. [NeurIPS2025] The official implementation of MindOmni: Unleashing Reasoning Generation in Vision Language Models with RGPO

UniWorld. UniWorld: High-Resolution Semantic Encoders for Unified Visual Understanding and Generation

WISE. [ICML 2026🔥] WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generation

mega-sam. Code for the project "MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos"

HunyuanVideo-Avatar. Python

Jenga. [NeurIPS 2025] Training-Free Efficient Video Generation via Dynamic Token Carving

Bagel. Open-source unified multimodal model

CameraBench. [NeurIPS 2025 Spotlight] Towards Understanding Camera Motions in Any Video

Qwen3. Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.

Awesome-From-Video-Generation-to-World-Model. A list of works on video generation towards world model

Step1X-Edit. A SOTA open-source image editing model, which aims to provide comparable performance against the closed-source models like GPT-4o and Gemini 2 Flash.

UniAnimate. Code for SCIS-2025 Paper "UniAnimate: Taming Unified Video Diffusion Models for Consistent Human Image Animation".

Awesome-Controllable-Video-Generation. [ArXiv 2025] A survey about controllable video generation: This repo is the official awesome of "Controllable video generation: A survey"

MAGI-1. MAGI-1: Autoregressive Video Generation at Scale

FramePack. Lets make video diffusion practical!

awesome-gpt4o-images. Awesome curated collection of images and prompts generated by GPT-4o and gpt-image-1. Explore AI generated visuals created with ChatGPT and Sora, showcasing OpenAI’s advanced image generation capabilities.

metamorph. Code for MetaMorph Multimodal Understanding and Generation via Instruction Tuning

Geo4D. [ICCV 2025 Highlight] Geo4D: Leveraging Video Generators for Geometric 4D Scene Reconstruction

Emu3. Next-Token Prediction is All You Need

SkyReels-A2. SkyReels-A2: Compose anything in video diffusion transformers

GPT-ImgEval. GPT-ImgEval: Evaluating GPT-4o’s state-of-the-art image generation capabilities

Janus. Janus-Series: Unified Multimodal Understanding and Generation Models

AURORA. Code and data for the paper: Learning Action and Reasoning-Centric Image Editing from Videos and Simulation

MetaCLIP. NeurIPS 2025 Spotlight; ICLR2024 Spotlight; CVPR 2024; EMNLP 2024

Lumina-mGPT-2.0. Lumina-mGPT 2.0: Stand-Alone AutoRegressive Image Modeling

DreamEngine. Multimodal Representation Alignment for Image Generation: Text-Image Interleaved Control Is Easier Than You Think!

mar. PyTorch implementation of MAR+DiffLoss https://arxiv.org/abs/2406.11838

NOVA. [ICLR 2025] Autoregressive Video Generation without Vector Quantization

Qwen2.5-Omni. Qwen2.5-Omni is an end-to-end multimodal model by Qwen team at Alibaba Cloud, capable of understanding text, audio, vision, video, and performing real-time speech generation.

DiffMoE. [Arxiv 2025] Official PyTorch implementation of DiffMoE, TC-DiT, EC-DiT and Dense DiT

ddpo-pytorch. DDPO for finetuning diffusion models, implemented in PyTorch with LoRA support

Awesome-MCoT. Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey

GoT. Official repository of "GoT: Unleashing Reasoning Capability of Multimodal Large Language Model for Visual Generation and Editing"

TeaCache. Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model

ACE_plus. Python

LBM. Latent Bridge Matching for Fast Image-to-Image Translation (ICCV 2025 Highlight)

SpargeAttn. [ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

HunyuanVideo-I2V. HunyuanVideo-I2V: A Customizable Image-to-Video Model based on HunyuanVideo

Visual-RFT. Official repository of 'Visual-RFT: Visual Reinforcement Fine-Tuning' & 'Visual-ARFT: Visual Agentic Reinforcement Fine-Tuning'’

MovieBench. [CVPR 2025] A Hierarchical Movie Level Dataset for Long Video Generation

Awesome-World-Model. Collect some World Models for Autonomous Driving (and Robotic, etc.) papers.

EDG. [CVPR2025] Extrapolating and Decoupling Image-to-Video Generation Models: Motion Modeling is Easier Than You Think

LightningDiT. [CVPR 2025 Oral] Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models

CameraViewer. A lightweight tool for camera pose visualization

DiT-Extrapolation. Official implementation for "RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers" (ICML 2025) , UltraViCo (ICLR 2026) and UltraImage

Wan2.1. Wan: Open and Advanced Large-Scale Video Generative Models

TrajectoryCrafter. [ICCV 2025, Oral] TrajectoryCrafter: Redirecting Camera Trajectory for Monocular Videos via Diffusion Models

Step-Video-T2V. Python

HunyuanVideoGP. HunyuanVideo GP: Large Video Generation Model - GPU Poor version

Show-o. [ICLR & NeurIPS 2025] Repository for Show-o series, One Single Transformer to Unify Multimodal Understanding and Generation.

FlashVideo. [AAAI-2026]FlashVideo: Flowing Fidelity to Detail for Efficient High-Resolution Video Generation

goku. [CVPR2025 Highlight] Video Generation Foundation Models: https://saiyan-world.github.io/goku/

Go-with-the-Flow. The official implementation of CVPR'25 Oral paper "Go-with-the-Flow: Motion-Controllable Video Diffusion Models Using Real-Time Warped Noise"

TransPixeler. CVPR2025

MagicMirror. [ICCV 2025] MagicMirror: ID-Preserved Video Generation in Video Diffusion Transformers

CamI2V. official repo of paper for "CamI2V: Camera-Controlled Image-to-Video Diffusion Model"

Tora. [CVPR'25]Tora: Trajectory-oriented Diffusion Transformer for Video Generation

LLMs_from_scratch. Learning records for building a large language model from scratch

VideoVAEPlus. [ICCV 2025] VideoVAE+: Large Motion Video Autoencoding with Cross-modal Video VAE

LuminaBrush. Illumination Drawing Tools for Text-to-Image Diffusion Models

genesis-world. Simulation platform for general-purpose robotics & embodied AI learning.

Lyra. [ICCV 2025] Official Implementation for "Lyra: An Efficient and Speech-Centric Framework for Omni-Cognition"

FastVideo. A unified inference and post-training framework for accelerated video generation.

ditflow. Official PyTorch implementation - Video Motion Transfer with Diffusion Transformers

HunyuanVideo. HunyuanVideo: A Systematic Framework For Large Video Generation Model

LTX-Video. Official repository for LTX-Video

Reducio-VAE. Jupyter Notebook

samurai. Official repository of "SAMURAI: Adapting Segment Anything Model for Zero-Shot Visual Tracking with Motion-Aware Memory"

In-Context-LoRA. Official repository of In-Context LoRA for Diffusion Transformers

ml-depth-pro. Depth Pro: Sharp Monocular Metric Depth in Less Than a Second.

MoGe. [CVPR'25 Oral] MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision

finetrainers. Scalable and memory-optimized training of diffusion models

InstaFlow. :zap: InstaFlow! One-Step Stable Diffusion with Rectified Flow (ICLR 2024)

SG-I2V. This is the official implementation of SG-I2V: Self-Guided Trajectory Control in Image-to-Video Generation.

Cosmos-Tokenizer. A suite of image and video neural tokenizers

Ctrl-V. 👆Pytorch implementation of "Ctrl-V: Higher Fidelity Video Generation with Bounding-Box Controlled Object Motion"

yt-dlp. A feature-rich command-line audio/video downloader

gradio-image-prompter. Image Prompter for Gradio

VideoTuna. Let's finetune video generation models!

GroundingDINO. [ECCV 2024] Official implementation of the paper "Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection"

Tracking-Anything-with-DEVA. [ICCV 2023] Tracking Anything with Decoupled Video Segmentation

Framer. [ICLR'25] Official PyTorch implementation of "Framer: Interactive Frame Interpolation".

mochi. The best OSS video generation models, created by Genmo

conditional-motion-propagation. Code for our CVPR 2019 work.

Allegro. Allegro is a powerful text-to-video model that generates high-quality videos up to 6 seconds at 15 FPS and 720p resolution from simple text input.

Sana. SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

DragNUWA.

torch-dct. DCT (discrete cosine transform) functions for pytorch

MovieGenBench. Movie Gen Bench - two media generation evaluation benchmarks released with Meta Movie Gen

LivePhoto. Official implementations for paper: LivePhoto: Real Image Animation with Text-guided Motion Control

ControlAR. [ICLR 2025] ControlAR: Controllable Image Generation with Autoregressive Models

monst3r. Official Implementation of paper "MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion"

Pyramid-Flow. [ICLR 2025] Pyramidal Flow Matching for Efficient Video Generative Modeling

Director3D. Code for "Director3D: Real-world Camera Trajectory and 3D Scene Generation from Text" (NeurIPS 2024).

VideoX-Fun. 📹 A more flexible framework that can generate videos at any resolution and creates videos from images.

DrawingSpinUp. (SIGGRAPH Asia 2024) This is the official PyTorch implementation of SIGGRAPH Asia 2024 paper: DrawingSpinUp: 3D Animation from Single Character Drawings

WritingAIPaper. Writing AI Conference Papers: A Handbook for Beginners

xDiT. xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism

Vertical. [ECCV2022] Official implementation of the paper "View Vertically: A Hierarchical Network for Trajectory Prediction via Fourier Spectrums"

ToonCrafter-for-windows. Python

generative-image-dynamics. Jupyter Notebook

AudioToText. Transcribe and translate audio to text using Whisper and DeepL.

taesd. Tiny AutoEncoder for Stable Diffusion (and other image models)

FollowYourCanvas. [AAAI 2025] Follow-Your-Canvas: This repo is the official implementation of "Follow-Your-Canvas: Higher-Resolution Video Outpainting with Extensive Content Generation"