Yukun Huang

rcm. rCM & Causal-rCM: Leading and Unified Algorithms/Infrastructures for Bidirectional/Autoregressive Video Diffusion Distillation at Scale

OmniX. Official implementation of "OmniX: From Unified Panoramic Generation and Perception to Graphics-Ready 3D Scenes".

flashdreams. high-performance inference and serving library for interactive autoregressive video and world models

taehv. Tiny AutoEncoder for Hunyuan Video (and other video models)

SDPO. Reinforcement Learning via Self-Distillation (SDPO)

PhysForge. [ICML 2026] PhysForge: Generating Physics-Grounded 3D Assets for Interactive Virtual World

DeepForcing. [ICML 2026] Official implementation of "Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression"

video2dataset. Easily create large video dataset from video urls

video-dataset-scripts. Collection of scripts to build small-scale datasets for fine-tuning video generation models.

clash-verge-rev. A modern GUI client based on Tauri, designed to run in Windows, macOS and Linux for tailored proxy experience

Infinite-Forcing. Infinite-Forcing: Towards Infinite-Long Video Generation

Wan2.2-TI2V-5B-Turbo. 4-steps distilled version of Wan2.2-TI2V-5B

Frequency-aware-Length-EXtension. official implementation for paper titled "Training-free Horizon Extension for Autoregressive Video Generation"

CubiD. [CVPR2026 Highlight] Cubic Discrete Diffusion: Discrete Visual Generation on High-Dimensional Representation Tokens https://arxiv.org/abs/2603.19232

Stable-Video-Infinity. [ICLR 26 Oral] Stable Video Infinity: Infinite-Length Video Generation with Error Recycling

DreamCube. [ICCV 2025] Official implementation of the paper "DreamCube: 3D Panorama Generation via Multi-plane Synchronization".

Person-Attribute-Recognition-MarketDuke. A simple baseline implemented in PyTorch for pedestrian attribute recognition task, evaluating on Market-1501 and DukeMTMC-reID dataset.

Math-VR-CodePlot-CoT. Math-VR Benchmark & CodePlot-CoT: Mathematical Visual Reasoning by Thinking with Code-Driven Images

OmniPart. [SIGGRAPH Asia 2025] OmniPart: Part-Aware 3D Generation with Semantic Decoupling and Structural Cohesion

MultiPanoWise. MULTIPANOWISE: HOLISTIC DEEP ARCHITECTURE FOR MULTI-TASK DENSE PREDICTION FROM A SINGLE PANORAMIC IMAGE

Intrinsic. Repo for the papers "Intrinsic Image Decomposition via Ordinal Shading" (TOG 2023) and "Colorful Diffuse Intrinsic Image Decomposition in the Wild" (TOG 2024)

prism-bench. This is the official repository for the paper "FLUX-Reason-6M & PRISM-Bench: A Million-Scale Text-to-Image Reasoning Dataset and Comprehensive Benchmark"

T2I-ReasonBench. T2I-ReasonBench: Benchmarking Reasoning-Informed Text-to-Image Generation

Matrix-3D. Generate large-scale explorable 3D scenes with high-quality panorama videos from a single image or text prompt.

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

bpy-renderer. Python package for rendering 3D scenes and animations using blender.

OST-Bench. [NeurIPS 2025] OST-Bench: Evaluating the Capabilities of MLLMs in Online Spatio-temporal Scene Understanding

StreamVLN. [ICRA 2026] Official implementation of the paper: "StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modeling"

GRPO-CARE. [ACL2026 Findings] GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning

Anime-Shooter. Python

GoT-R1. [ICLR26] GoT-R1: Unleashing Reasoning Capability of MLLM for Visual Generation with Reinforcement Learning

WorldGen. 🌍 WorldGen - Generate Any 3D Scene in Seconds

T2I-Personalization-with-AR.

FramePack. Lets make video diffusion practical!

GigaTok. [ICCV 2025] Official repo for "GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation"

HoloPart. HoloPart: Generative 3D Part Amodal Segmentation

GameFactory. [ICCV 2025] GameFactory: Creating New Games with Generative Interactive Videos

HEAL-SWIN. Reference implementation of the spherical vision transformer HEAL-SWIN

Deformable-Radial-Kernel-Splatting. [CVPR 2025] Code for Deformable Radial Kernel Splatting

Imagine360. [Neurips 2025] Imagine360: Immersive 360 Video Generation from Perspective Anchor

Moto. [ICCV2025 Oral] Latent Motion Token as the Bridging Language for Learning Robot Manipulation from Videos

GenMAC. [AAAI 2026] GenMAC for Compositional Text-to-Video Generation

iGibson. A Simulation Environment to train Robots in Large Realistic Interactive Scenes

GaussianObject. GaussianObject: High-Quality 3D Object Reconstruction from Four Views with Gaussian Splatting (SIGGRAPH Asia 2024, TOG)

DreamCraft3D_Plus. [TPAMI 2025] DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model

PUMA. Empowering Unified MLLM with Multi-granular Visual Generation

equilib. 🌎→🗾Equirectangular (360/panoramic) image processing library for Python with minimal dependencies only using Numpy and PyTorch

py360convert. Python implementation of convertion between equirectangular, cubemap and perspective. (equirect2cube, cube2equirect, equirect2perspec)

Perspective-and-Equirectangular. A tool to transfer Perspective between Equirectangular

Lotus. Official implementation of "Lotus: Diffusion-based Visual Foundation Model for High-quality Dense Prediction"

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

transfusion-pytorch. Pytorch implementation of Transfusion, "Predict the Next Token and Diffuse Images with One Multi-Modal Model", from MetaAI

ViewCrafter. [TPAMI 2025] ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis

EscherNet. [CVPR2024 Oral] EscherNet: A Generative Model for Scalable View Synthesis

PanFusion. 🍳 [CVPR'24 Highlight] Pytorch implementation of "Taming Stable Diffusion for Text to 360° Panorama Image Generation"

splat. WebGL 3D Gaussian Splat Viewer

LayerPano3D. [SIGGRAPH 2025] LayerPano3D: Layered 3D Panorama for Hyper-Immersive Scene Generation"

pixelsplat. [CVPR 2024 Oral, Best Paper Runner-Up] Code for "pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction" by David Charatan, Sizhe Lester Li, Andrea Tagliasacchi, and Vincent Sitzmann

DressCode. DressCode: Autoregressively Sewing and Generating Garments from Text Guidance.

DDBM. Python

dust3r. DUSt3R: Geometric 3D Vision Made Easy

GaussianCube. [NeurIPS 2024] GaussianCube: A Structured and Explicit Radiance Representation for 3D Generative Modeling

Physics3D. Official implementation of Physics3D: Learning Physical Properties of 3D Gaussians via Video Diffusion

TalkSHOW. This is the official repository for TalkSHOW: Generating Holistic 3D Human Motion from Speech [CVPR2023].

SHOW. This is the codebase for SHOW in Generating Holistic 3D Human Motion from Speech [CVPR2023],

3D-Gaussian-Splatting-Papers. 3D高斯论文，持续更新，欢迎交流讨论。

particle-sfm. ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wild. ECCV 2022.

mvdream_diffusers. A unified diffusers implementation for MVDream and ImageDream

TripoSR. TripoSR: Fast 3D Object Reconstruction from a Single Image

LayerDiffuse. Transparent Image Layer Diffusion using Latent Transparency

Deformable-3D-Gaussians. [CVPR 2024] Official implementation of "Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction"

Relightable3DGaussian. [ECCV2024] Relightable 3D Gaussian: Real-time Point Cloud Relighting with BRDF Decomposition and Ray Tracing

Open-Sora-Plan. This project aim to reproduce Sora (Open AI T2V model), we wish the open source community contribute to this project.

LucidDreamer. Official implementation of "LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching"

gaussian_splatting_notes. A detailed formulae explanation on gaussian splatting

gaussian-mesh-splatting. Official implementation of "Mesh-Based Adapting and Modification of Gaussian Splatting"

warp. A Python framework for GPU-accelerated simulation, robotics, and machine learning.

gala. [CVPR 2024] GALA: Generating Animatable Layered Assets from a Single Scan

GaussianAvatar. [CVPR 2024] The official repo for "GaussianAvatar: Towards Realistic Human Avatar Modeling from a Single Video via Animatable 3D Gaussians"

Awesome-3D-AIGC. A curated list of papers and open-source resources focused on 3D AIGC.

GShell. Official implentation of "Ghost on the Shell: An Expressive Representation of General 3D Shapes" (ICLR 2024 Oral)

Awesome-4D-Generation. An organized list of academic papers focused on the topic of 4D Generation. If you have any additions or suggestions, feel free to contribute.

LayersNet_ICCV2023. Python

Open-AnimateAnyone. Unofficial Implementation of Animate Anyone

gradsim. Differentiable simulation for system identification and visuomotor control

AnimateAnyone. Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation

threestudio-dreamwaltz. DreamWaltz extension of threestudio.

SuGaR. [CVPR 2024] Official PyTorch implementation of SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering

PhysGaussian. [CVPR 2024 Highlight] PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics

FlexiCubes. Python

humannorm. CVPR 2024: The official implementation of HumanNorm

HumanGaussian. [CVPR 2024 Highlight] Code for "HumanGaussian: Text-Driven 3D Human Generation with Gaussian Splatting"

awesome-3D-gaussian-splatting. Curated list of papers and resources focused on 3D Gaussian Splatting, intended to keep pace with the anticipated surge of research in the coming months.

GART. GART: Gaussian Articulated Template Models

DreamComposer. [CVPR 2024] DreamComposer: Controllable 3D Object Generation via Multi-View Conditions

MVDream. Multi-view Diffusion for 3D Generation

gsgen. [CVPR 2024] Text-to-3D using Gaussian Splatting

dreamgaussian.github.io. Project Page for Generative Gaussian Splatting for Efficient 3D Content Creation

instant-nsr-pl. Neural Surface reconstruction based on Instant-NGP. Efficient and customizable boilerplate for your research projects. Train NeuS in 10min!

IT3D-text-to-3D. [AAAI'2024] IT3D: Improved Text-to-3D Generation with Explicit View Synthesis

buddi. Our model BUDDI learns the joint distribution of interacting people

Hi3D. From Geometry to Texture: A Hierarchical Framework for Efficient Text-to-3D Generation

3D-deformable-attention. [ICCV 2023] Official implementation of the paper "DFA3D: 3D Deformable Attention For 2D-to-3D Feature Lifting"

hana. Implementation and checkpoints of Imagen, Google's text-to-image synthesis neural network, in Pytorch

Lora-for-Diffusers. The most easy-to-understand tutorial for using LoRA (Low-Rank Adaptation) within diffusers framework for AI Generation Researchers🔥

dreambooth3d.github.io. HTML

MoPA. An Blender addon uses ROMP to extract human's 3D poses from image, video or webcam and drive your own 3D character.

4D-Humans. 4DHumans: Reconstructing and Tracking Humans with Transformers

threestudio. A unified framework for 3D content generation.

HumanSD. [ICCV 2023] The official implementation of paper "HumanSD: A Native Skeleton-Guided Diffusion Model for Human Image Generation"

human_body_prior. VPoser: Variational Human Pose Prior

Pose_to_SMPL. A tool to fit SMPL parameters from 3D-pose datasets that contain key-points of human body.

SMPL. NumPy, TensorFlow and PyTorch implementation of human body SMPL model and infant body SMIL model.

Minimal-IK. Solving SMPL/MANO parameters from keypoint coordinates.

smplx. SMPL-X

ustcthesis. LaTeX template for USTC thesis

P2-weighting. CVPR 2022

torch-ngp. A pytorch CUDA extension implementation of instant-ngp (sdf and nerf), with a GUI.

sjc. Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation (CVPR 2023)

learnit. Jupyter Notebook

latent-nerf. Official Implementation for "Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures"

stable-dreamfusion. Text-to-3D & Image-to-3D & Mesh Exportation with NeRF + Diffusion.

3DCommonCorruptions. 3D Common Corruptions and Data Augmentation, CVPR 2022 [Oral]

MiDaS. Code for robust monocular depth estimation described in "Ranftl et. al., Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer, TPAMI 2022"

LUPerson-NL. Large-Scale Pre-training for Person Re-identification with Noisy Labels (LUPerson-NL)

denoising-diffusion-pytorch. Implementation of Denoising Diffusion Probabilistic Model in Pytorch

MITML. Python

HITSZ-VCM-data.

deepchange. ICCV 2023, project page of the paper "DeepChange: A Long-term Person Re-identification Benchmark"

MINE-Mutual-Information-Neural-Estimation-. A pytorch implementation of MINE(Mutual Information Neural Estimation)

CIL-ReID. Benchmarks for Corruption Invariant Person Re-identification. [NeurIPS 2021 Track on Datasets and Benchmarks]

Barbershop. Barbershop: GAN-based Image Compositing using Segmentation Masks (SIGGRAPH Asia 2021)

EnlightenGAN-inference. ONNX Inference library for EnlightenGAN

DSFD_pytorch. pytorch 实现的DSFD， 更高效更简洁

stargan-v2. StarGAN v2 - Official PyTorch Implementation (CVPR 2020)

SPADE. Semantic Image Synthesis with SPADE

Neural-Pose-Transfer. Neural Pose Transfer by Spatially Adaptive Instance Normalization. In CVPR 2020

NAD. The project page of paper: Architecture disentanglement for deep neural networks [ICCV 2021, oral]

VERI-Wild. A Large-Scale Dataset for Vehicle Re-Identification in the Wild

RainDirection-and-Real3000-Dataset. This repository contains an introduction and links to RainDirection and Real3000 datasets for paper Unpaired Learning for Deep Image Deraining with Rain Direction Regularizer, Authors:Yang Liu, Ziyu Yue ,Jinshan Pan, Zhixun Su, Dalian University of Technology and Nanjing University of Science and Technology

bolei_awesome_posters. CVPR and NeurIPS poster examples and templates

Pytorch_Retinaface. Retinaface get 80.99% in widerface hard val using mobilenet0.25.

DAGL. Dynamic Attentive Graph Learning for Image Restoration, ICCV2021 [PyTorch Code]

robustness. Corruption and Perturbation Robustness (ICLR 2019)

CVFPaperHelper. Automatically download multiple papers by keywords in CVPR

Generalization-Causality. 关于domain generalization，domain adaptation，causality，robutness，prompt，optimization，generative model各式各样研究的阅读笔记

yt-visionseed-sdk-python. python-sdk for VisionSeed - a camera module with AI ability

deep-learning-papers-translation. 深度学习论文翻译，包括分类论文，检测论文等

TransReID. [ICCV-2021] TransReID: Transformer-based Object Re-Identification

UAV-Human. [CVPR2021] UAV-Human: A Large Benchmark for Human Behavior Understanding with Unmanned Aerial Vehicles

pytorch-YOLOv4. PyTorch ,ONNX and TensorRT implementation of YOLOv4

faster-rcnn.pytorch. A faster pytorch implementation of faster r-cnn

maskrcnn-benchmark. Fast, modular reference implementation of Instance Segmentation and Object Detection algorithms in PyTorch.

continual-learning. PyTorch implementation of various methods for continual learning (XdG, EWC, SI, LwF, FROMP, DGR, BI-R, ER, A-GEM, iCaRL, Generative Classifier) in three different scenarios.

cluster-contrast-reid. Python

BIMEF. Code and data for the research paper "A Bio-Inspired Multi-Exposure Fusion Framework for Low-light Image Enhancement" (Submitted to IEEE Transactions on Cybernetics)

Lighting-the-Darkness-in-the-Deep-Learning-Era-Open.

ISM-ReID. The source code of paper “HAZY RE-ID: AN INTERFERENCE SUPPRESSION MODEL FOR DOMAIN ADAPTATION PERSON RE-IDENTIFICATION UNDER INCLEMENT WEATHER CONDITION”.

CVPR2024-Paper-Code-Interpretation. cvpr2024/cvpr2023/cvpr2022/cvpr2021/cvpr2020/cvpr2019/cvpr2018/cvpr2017 论文/代码/解读/直播合集，极市团队整理

VIDIT. (CVPR21/ECCV20 Workshops) Official repo: Virtual Image Dataset for Illumination Transfer (VIDIT)

HLA-Face-Code. Code for HLA-Face: Joint High-Low Adaptation for Low Light Face Detection (CVPR21)

deep-person-reid. Torchreid: Deep learning person re-identification in PyTorch.

pytorch-OpCounter. Count the MACs / FLOPs of your PyTorch model.

LatentGNN-V1-PyTorch. Implementation of “LatentGNN: Learning Efficient Non-local Relations for Visual Recognition”(ICML 2019)

Adversarial-attack-on-Person-ReID-With-Deep-Mis-Ranking. This is a pytorch implementation of the CVPR2020 paper: Transferable, Controllable, and Inconspicuous Adversarial Attacks on Person Re-identification With Deep Mis-Ranking

fast-reid. SOTA Re-identification Methods and Toolbox

GitHub520. :kissing_heart: 让你“爱”上 GitHub，解决访问时图裂、加载慢的问题。（无需安装）

FGSM. Simple pytorch implementation of FGSM and I-FGSM

pytorch-classification-advprop. Python

Deep-adversarial-decomposition. Pytorch implementation of the paper: "A Unified Framework for Separating Superimposed Images", in CVPR 2020.

ISDA-for-Deep-Networks. An efficient implicit semantic augmentation method, complementary to existing non-semantic techniques.

OWM. Code for Continual Learning of Context-dependent Processing in Neural Networks

EnlightenGAN. [IEEE TIP] "EnlightenGAN: Deep Light Enhancement without Paired Supervision" by Yifan Jiang, Xinyu Gong, Ding Liu, Yu Cheng, Chen Fang, Xiaohui Shen, Jianchao Yang, Pan Zhou, Zhangyang Wang

Predict-Cluster. Repository for PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition

nni. An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.

ToDayGAN. Python

Gumbel_Softmax_VAE. PyTorch implementation of a Variational Autoencoder with Gumbel-Softmax Distribution

reid-strong-baseline. Bag of Tricks and A Strong Baseline for Deep Person Re-identification

Pedestrian-Attribute-Recognition-Paper-List. [Survey@Pattern Recognition] Paper list on Pedestrian Attribute Recognition (PAR) and related tasks (Pattern Recognition 2021)

leetcode. LeetCode题解，151道题完整版。

AI-Job-Notes. AI算法岗求职攻略（涵盖准备攻略、刷题指南、内推和AI公司清单等资料）

Pytorch-UNet. PyTorch implementation of the U-Net for image semantic segmentation with high quality images

autosub. [NO LONGER MAINTAINED] Command-line utility for auto-generating subtitles for any video file

person-re-ranking. Person Re-ranking (CVPR 2017)

Exclusively-Dark-Image-Dataset. Exclusively Dark (ExDARK) dataset which to the best of our knowledge, is the largest collection of low-light images taken in very low-light environments to twilight (i.e 10 different conditions) to-date with image class and object level annotations.

awesome-reid-dataset. Collection of public available person re-identification datasets

pytorch-colors. Small utility for color space conversions of PyTorch Tensors and Variables

pytorch-ssim. pytorch structural similarity (SSIM) loss

Person_reID_baseline_pytorch. :bouncing_ball_person: Pytorch ReID: A tiny, friendly, strong pytorch implement of person re-id / vehicle re-id baseline. Tutorial 👉https://github.com/layumi/Person_reID_baseline_pytorch/tree/master/tutorial

ipv6-hosts. Fork of https://code.google.com/archive/p/ipv6-hosts/, focusing on automation

deep-learning-model-convertor. The convertor/conversion of deep learning models for different deep learning frameworks/softwares.

ustcthesis. Moved to USTCTUG on Github中国科学技术大学学位论文通用LaTeX模板. LaTeX templates for UNDERGRADUATE and GRADUATE STUDENTS of USTC. Backup mirror at GitLab (sponsored by LUG@USTC) : https://gitlab.lug.ustc.edu.cn/ywg/ustcthesis