Qiqi Hou

OpenSpec. Spec-driven development (SDD) for AI coding assistants.

SDPO. Reinforcement Learning via Self-Distillation (SDPO)

Motus. Official code of Motus: A Unified Latent Action World Model

lingbot-va. [RSS 2026] Causal video-action world model for generalist robot control

SimpleVLA-RL. [ICLR 2026] SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

GDPO. Official implementation of GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization

OpenTinker. OpenTinker is an RL-as-a-Service infrastructure for foundation models

mup. maximal update parametrization (µP)

PaCoRe. PaCoRe: Learning to Scale Test-Time Compute with Parallel Coordinated Reasoning

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

Depth-Anything-3. Depth Anything 3

lejepa. Python

cosmos-predict2.5. Cosmos-Predict2.5, the latest version of the Cosmos World Foundation Models (WFMs) family, specialized for simulating and predicting the future state of the world in the form of video.

map-anything. MapAnything: Universal Feed-Forward Metric 3D Reconstruction

remake. やり直すんだ。そして、次はうまくやる。

sort-free-gs. Re-implementation of Sort-Free Gaussian Splatting via Weighted Sum Rendering

CUT3R. Official implementation of Continuous 3D Perception Model with Persistent State

native-sparse-attention-triton. Efficient triton implementation of Native Sparse Attention.

fast3r. [CVPR 2025] Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass

prope. Cameras as Relative Positional Encoding

AnySplat. [SIGGRAPH Asia 2025 (ACM TOG)] AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Views

Direct3D-S2. [NeurIPS 2025] Direct3D‑S2: Gigascale 3D Generation Made Easy with Spatial Sparse Attention

Long-LRM. Self-reimplemented version of Long-LRM.

LVSM. [ICLR 2025 Oral] Official code for "LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias"

ttt-video-dit. Official PyTorch implementation of One-Minute Video Generation with Test-Time Training

vggt. [CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer

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

GPT-SoVITS. 1 min voice data can also be used to train a good TTS model! (few shot voice cloning)

R1-V. Witness the aha moment of VLM with less than $3.

PIKE-RAG. PIKE-RAG: sPecIalized KnowledgE and Rationale Augmented Generation

ChatDev. ChatDev 2.0: Dev All through LLM-powered Multi-Agent Collaboration

LLaVA. [NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

lmms-eval. One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

TokenFlow. [CVPR 2025] 🔥 Official impl. of "TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation".

LLM-Reverse-Curriculum-RL. Implementation of the ICML 2024 paper "Training Large Language Models for Reasoning through Reverse Curriculum Reinforcement Learning" presented by Zhiheng Xi et al.

gsplat-website. Contains website data for https://gsplat.tech

splat. WebGL 3D Gaussian Splat Viewer

NoPoSplat. [ICLR'25 Oral] No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images

xformers. Hackable and optimized Transformers building blocks, supporting a composable construction.

Compact-3DGS. The official repository of Compact 3D Gaussian Representation for Radiance Field

gsplat. CUDA accelerated rasterization of gaussian splatting

timeful.app. Timeful (formerly Schej) is a scheduling platform helps you find the best time for a group to meet. It is a free availability poll that is easy to use and integrates with your calendar.

slang-gaussian-rasterization. Slang

clean-fid. PyTorch - FID calculation with proper image resizing and quantization steps [CVPR 2022]

dgm-eval. Codebase for evaluation of deep generative models as presented in Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models

DNGaussian. [CVPR'24] DNGaussian: Optimizing Sparse-View 3D Gaussian Radiance Fields with Global-Local Depth Normalization

multinerf. A Code Release for Mip-NeRF 360, Ref-NeRF, and RawNeRF

torch-splatting. A pure pytorch implementation of 3D gaussian Splatting

BlenderNeRF. Easy NeRF synthetic dataset creation within Blender

fwdgrad. Implementation of "Gradients without backpropagation" paper (https://arxiv.org/abs/2202.08587) using functorch

generative-compression. TensorFlow Implementation of Generative Adversarial Networks for Extreme Learned Image Compression

SpacetimeGaussians. [CVPR 2024] Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesis

generative-models. Generative Models by Stability AI

Dynamic3DGaussians. Python

slang. Making it easier to work with shaders

NeuRBF. Python

gaussian-splatting. Original reference implementation of "3D Gaussian Splatting for Real-Time Radiance Field Rendering"

mitsuba3. Mitsuba 3: A Retargetable Forward and Inverse Renderer

R-FID-Robustness-of-Quality-Measures-for-GANs. This is the official repo for the ECCV paper: "On the Robustness of Quality Measures for GANs

distill-sd. Segmind Distilled diffusion

crunch. Advanced DXTc texture compression and transcoding library

BK-SDM. A Compressed Stable Diffusion for Efficient Text-to-Image Generation [ECCV'24]

aimet-model-zoo. Python

RVRT. Recurrent Video Restoration Transformer with Guided Deformable Attention (NeurlPS2022, official repository)

BasicVSR_PlusPlus. Official repository of "BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignment"

monodepth2. [ICCV 2019] Monocular depth estimation from a single image

EMA-VFI. [CVPR 2023] Extracting Motion and Appearance via Inter-Frame Attention for Efficient Video Frame Interpolatio

PCWNet. Python

PhoenixGo. Go AI program which implements the AlphaGo Zero paper

AoT_Dataset. CVPR18: Learning and Using the Arrow of Time

DCVC. Deep Contextual Video Compression

RealtimeStereo. Attention-Aware Feature Aggregation for Real-time Stereo Matching on Edge Devices (ACCV, 2020)

MC-Denoising-via-Auxiliary-Feature-Guided-Self-Attention. Official implementation of MC Denoising via Auxiliary Feature Guided Self-Attention (SIGGRAPH Asia 2021 paper)

vision_blender. A Blender addon for generating synthetic ground truth data for Computer Vision applications

pygta5. Explorations of Using Python to play Grand Theft Auto 5.

Awesome-Image-Harmonization. A curated list of papers, code and resources pertaining to image harmonization.

SimSwap. An arbitrary face-swapping framework on images and videos with one single trained model!

faceswap. Deepfakes Software For All

stereo-magnification. Code accompanying the SIGGRAPH 2018 paper "Stereo Magnification: Learning View Synthesis using Multiplane Images"

blender-cli-rendering. Blender Python scripts for rendering images directly from command-line interface

nvdiffrec. Official code for the CVPR 2022 (oral) paper "Extracting Triangular 3D Models, Materials, and Lighting From Images".

AirSim. Open source simulator for autonomous vehicles built on Unreal Engine / Unity, from Microsoft AI & Research

VSFA. [official] Quality Assessment of In-the-Wild Videos (ACM MM 2019)

DVQA. Deep learning-based Video Quality Assessment

tungsten. High performance physically based renderer in C++11

AI-Paper-Collector. MLNLP社区用来更好进行论文搜索的工具。Fully-automated scripts for collecting AI-related papers

face-parsing.PyTorch. Using modified BiSeNet for face parsing in PyTorch

face-seg. Semantic segmentation for hair, face and background

hair-segmentation. hair segmentation in mobile device

renderdoc_for_game_data. Access data with annotations from GTA5 using renderdoc

Video-Frame-Interpolation-Transformer. Python

HCFlow. Official PyTorch code for Hierarchical Conditional Flow: A Unified Framework for Image Super-Resolution and Image Rescaling (HCFlow, ICCV2021)

TensoRF. [ECCV 2022] Tensorial Radiance Fields, a novel approach to model and reconstruct radiance fields

unrealrox-plus. Upgraded version of UnrealROX, as plugin

gamehook. C++

FFHQ-Aging-Dataset. FFHQ-Aging Dataset

ExtraNet. ExtraNet: Real-time Extrapolated Rendering for Low-latency Temporal Supersampling

revisiting-sepconv. an implementation of Revisiting Adaptive Convolutions for Video Frame Interpolation using PyTorch

Context-Aware-Matting. The project is the inference implementation of our ICCV2019 paper "Context-aware Image Matting for Simultaneous Foreground and Alpha Estimation"

instant-ngp. Instant neural graphics primitives: lightning fast NeRF and more

raspiquickcv2. Install opencv in minutes on raspbian

SwinIR. SwinIR: Image Restoration Using Swin Transformer (official repository)

ml-cvnets. CVNets: A library for training computer vision networks

kilonerf. Code for KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs

hyperstyle. Official Implementation for "HyperStyle: StyleGAN Inversion with HyperNetworks for Real Image Editing" (CVPR 2022) https://arxiv.org/abs/2111.15666

SOAT. Official PyTorch repo for StyleGAN of All Trades: Image Manipulation with Only Pretrained StyleGAN.

SRCNN-pytorch. PyTorch implementation of Image Super-Resolution Using Deep Convolutional Networks (ECCV 2014)

photometric_optimization. Photometric optimization code for creating the FLAME texture space and other applications

FastFlowNet. FastFlowNet: A Lightweight Network for Fast Optical Flow Estimation (ICRA 2021)

VideoPose3D. Efficient 3D human pose estimation in video using 2D keypoint trajectories

PerceptualSimilarity. LPIPS metric. pip install lpips

UniPose. We propose UniPose, a unified framework for human pose estimation, based on our “Waterfall” Atrous Spatial Pooling architecture, that achieves state-of-art-results on several pose estimation metrics. Current pose estimation methods utilizing standard CNN architectures heavily rely on statistical postprocessing or predefined anchor poses for joint localization. UniPose incorporates contextual seg- mentation and joint localization to estimate the human pose in a single stage, with high accuracy, without relying on statistical postprocessing methods. The Waterfall module in UniPose leverages the efficiency of progressive filter- ing in the cascade architecture, while maintaining multi- scale fields-of-view comparable to spatial pyramid config- urations. Additionally, our method is extended to UniPose- LSTM for multi-frame processing and achieves state-of-the- art results for temporal pose estimation in Video. Our re- sults on multiple datasets demonstrate that UniPose, with a ResNet backbone and Waterfall module, is a robust and efficient architecture for pose estimation obtaining state-of- the-art results in single person pose detection for both sin- gle images and videos.

deep-head-pose. :fire::fire: Deep Learning Head Pose Estimation using PyTorch.

articulated-animation. Code for Motion Representations for Articulated Animation paper

first-order-model. This repository contains the source code for the paper First Order Motion Model for Image Animation

DeepV2D. Python

RAFT-3D. Python

MichiGAN. MichiGAN: Multi-Input-Conditioned Hair Image Generation for Portrait Editing (SIGGRAPH 2020)

expansion. Upgrading Optical Flow to 3D Scene Flow through Optical Expansion, CVPR 2020 (Oral).

Human-Segmentation-PyTorch. Human segmentation models, training/inference code, and trained weights, implemented in PyTorch

matting_human_datasets. 人像matting数据集，包含34427张图像和对应的matting结果图。

sbmc. Sample-based Monte Carlo Denoising using a Kernel-Splatting Network [Siggraph 2019]

BackgroundMattingV2. Real-Time High-Resolution Background Matting

deit. Official DeiT repository

vit-pytorch. Implementation of Vision Transformer, a simple way to achieve SOTA in vision classification with only a single transformer encoder, in Pytorch

oh-my-userscripts. 我的油猴脚本仓库

nerf. Code release for NeRF (Neural Radiance Fields)

kkndme_tianya. 天涯 kkndme 神贴聊房价

jonbarron.github.io. HTML

irr. Iterative Residual Refinement for Joint Optical Flow and Occlusion Estimation (CVPR 2019)

ScopeFlow. Dynamic Scene Scoping for Optical Flow (CVPR 2020)

BMBC. BMBC: Bilateral Motion Estimation with Bilateral Cost Volume for Video Interpolation, ECCV 2020

RAFT. Python

MaskFlownet-Pytorch. Pytorch implementation of MaskFlownet

FBA_Matting. Official repository for the paper F, B, Alpha Matting

MaskFlownet. [CVPR 2020, Oral] MaskFlownet: Asymmetric Feature Matching with Learnable Occlusion Mask

pytorch-randaugment. Unofficial PyTorch Reimplementation of RandAugment.

ARFlow. The official PyTorch implementation of the paper "Learning by Analogy: Reliable Supervision from Transformations for Unsupervised Optical Flow Estimation".

EDSR-PyTorch. PyTorch version of the paper 'Enhanced Deep Residual Networks for Single Image Super-Resolution' (CVPRW 2017)

SelFlow. SelFlow: Self-Supervised Learning of Optical Flow

closed-form-matting. Python implementation of A. Levin D. Lischinski and Y. Weiss. A Closed Form Solution to Natural Image Matting. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), June 2006, New York

softmax-splatting. an implementation of softmax splatting for differentiable forward warping using PyTorch

DRN. Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution

CAIN. Source code for AAAI 2020 paper "Channel Attention Is All You Need for Video Frame Interpolation"

pytorch_nms. CUDA implementation of NMS for PyTorch

PySceneDetect. :movie_camera: Python and OpenCV-based scene cut/transition detection program & library.

mogicians-manual. Flutter version【膜法指南】open source project

GCNet. GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond

CCNet. CCNet: Criss-Cross Attention for Semantic Segmentation (TPAMI 2020 & ICCV 2019).

evil-huawei. Evil Huawei - 华为作过的恶

blender-compile. (NOT MAINTIANED) docker environment for compiling blender 2.8

DALI. A GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.

tensorflow-densenet. Tensorflow-DenseNet with ImageNet Pretrained Models

OpenVINO-Custom-Layers. Tutorial for Using Custom Layers with OpenVINO (Intel Deep Learning Toolkit)

EDVR. Winning Solution in NTIRE19 Challenges on Video Restoration and Enhancement (CVPR19 Workshops) - Video Restoration with Enhanced Deformable Convolutional Networks. EDVR has been merged into BasicSR and this repo is a mirror of BasicSR.

knn-matting. Python implementation of KNN Matting, CVPR 2012 / TPAMI 2013 http://dingzeyu.li/projects/knn/

tpu. Reference models and tools for Cloud TPUs.

darts. Differentiable architecture search for convolutional and recurrent networks

awesome-architecture-search. A curated list of awesome architecture search resources

pytorch-mobilenet-v3. MobileNetV3 in pytorch and ImageNet pretrained models

HRNet-Semantic-Segmentation. The OCR approach is rephrased as Segmentation Transformer: https://arxiv.org/abs/1909.11065. This is an official implementation of semantic segmentation for HRNet. https://arxiv.org/abs/1908.07919

ranking. Learning to Rank in TensorFlow

tensorboardX. tensorboard for pytorch (and chainer, mxnet, numpy, ...)

Grid-Anchor-based-Image-Cropping. Project page of the CVPR2019 paper "Reliable and Efficient Image Cropping: A Grid Anchor based Approach"

ViewProposalNet. VPN for CVPR18

ViewEvaluationNet. VEN for cvpr2018

flickr-cropping-dataset. A dataset that includes photos downloaded from Flickr and annotations that indicates a local window representing a good composition.

AffinityBasedMattingToolbox. A collection of common affinity-based image matting and matte refinement algorithms.

view-finding-network. A deep ranking network that learns to find good compositions in a photograph.

PyTorch-YOLOv3. Minimal PyTorch implementation of YOLOv3

AlignedReID-Re-Production-Pytorch. Reproduce AlignedReID: Surpassing Human-Level Performance in Person Re-Identification, using Pytorch.

selfconsistency. Code for the paper: Fighting Fake News: Image Splice Detection via Learned Self-Consistency

DCGAN-tensorflow-slim. Implementation of DCGAN in TensorFlow-Slim

PWC-Net. PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume, CVPR 2018 (Oral)

semantic-segmentation-pytorch. Pytorch implementation for Semantic Segmentation/Scene Parsing on MIT ADE20K dataset

models. Models and examples built with TensorFlow

PyTorch-Encoding. A CV toolkit for my papers.

Detectron. FAIR's research platform for object detection research, implementing popular algorithms like Mask R-CNN and RetinaNet.

pytorch-semseg. Semantic Segmentation Architectures Implemented in PyTorch

tuio-mouse-driver. TUIO Mouse Driver

bottom-up-attention. Bottom-up attention model for image captioning and VQA, based on Faster R-CNN and Visual Genome

Deformable-ConvNets. Deformable Convolutional Networks + MST + Soft-NMS

improved_wgan_training. Code for reproducing experiments in "Improved Training of Wasserstein GANs"

XJTU-Thesis-Template. This is a LaTeX template for doctoral thesis of Xi`an Jiaotong University (XJTU). The platform is Windows + TexLive + XeLaTex.

WassersteinGAN. Python

fast-pixel-cnn. Speed up PixelCNN++ image generation by up to a 183 times

Faster-RCNN_TF. Faster-RCNN in Tensorflow

keras. Deep Learning for humans

deeplearning-papernotes. Summaries and notes on Deep Learning research papers

AdversarialNetsPapers. Awesome paper list with code about generative adversarial nets

WassersteinGAN.tensorflow. Tensorflow implementation of Wasserstein GAN - arxiv: https://arxiv.org/abs/1701.07875

tensorflow-zh. 谷歌全新开源人工智能系统TensorFlow官方文档中文版

iGAN. Interactive Image Generation via Generative Adversarial Networks

vae_tutorial. Caffe code to accompany my Tutorial on Variational Autoencoders

face-py-faster-rcnn. Face Detection with the Faster R-CNN

caffe. Caffe: a fast open framework for deep learning.

caffe. A fork of Caffe with OpenMPI-based Multi-GPU (mainly data parallel) support for action recognition and more. More documentation please see the original readme.

XX-Net. A proxy tool to bypass GFW.

YCM-Generator. Generates config files for YouCompleteMe (https://github.com/Valloric/YouCompleteMe)

spf13-vim. The ultimate vim distribution