Dusan Svilarkovic

IsaacLab. Unified framework for robot learning built on NVIDIA Isaac Sim

EgoScaler. [CVPR 2025 highlight] Generating 6DoF Object Manipulation Trajectories from Action Description in Egocentric Vision

Causal-Forcing. [ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++

vulture. Find dead Python code

tinyworlds. A minimal implementation of DeepMind's Genie world model

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

diso. Differentiable Iso-Surface Extraction Package (DISO)

Rex-Omni. [CVPR2026] Detect Anything via Next Point Prediction

foci. Fast, orientation-aware trajectory planning using a novel Gaussian overlap-based collision formulation, modeling both robot and environment as Gaussian Splat.

Pi3. [ICLR 2026] π^3: Permutation-Equivariant Visual Geometry Learning

mvtracker. [ICCV 2025 Oral] MVTracker: Multi-view 3D Point Tracking

LaRender. Python

system_prompts_leaks. Extracted system prompts from Anthropic - Claude Fable 5, Opus 4.8, Claude Code, Claude Design. OpenAI - ChatGPT GPT-5.6, Codex GPT-5.6, GPT-5.5. Google - Gemini 3.5 Flash, 3.1 Pro, Antigravity. xAI - Grok, Cursor, Copilot, VS Code, Perplexity, and more. Updated regularly.

STream3R. Dynamic 3D Foundation Model using Causal Transformer. [ICLR 2026]

YUME. The official code of Yume

slambook-en. The English version of 14 lectures on visual SLAM.

awesome-3d-point-cloud-denoising. A curated list of awesome 3D point cloud denoising papers

deeplearning_meshes_pointclouds. Code for SGP 2025 Graduate School tutorial "Deep Learning on Meshes and Point Clouds"

nice-slam. [CVPR'22] NICE-SLAM: Neural Implicit Scalable Encoding for SLAM

utils3d. A collection of useful functions for 3D vision & graphics research in Python.

TRELLIS. TRELLIS fork with additional memory handling.

sekai-codebase. [NeurIPS 2025] Sekai: A Video Dataset towards World Exploration

ntu-graduate-courses. Graduate-Level Courses (Electrical Engineering & Computer Science) at NTU, Singapore

GenNBV. [CVPR 2024] GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstruction

outformer. Clean Outputs from Language Models

AFLDM. [CVPR 2025 Oral] Alias-free Latent Diffusion Models (official implementation)

torch_packages_builder. Builder and index for PyTorch packages

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

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

SynCamMaster. [ICLR'25] SynCamMaster: Synchronizing Multi-Camera Video Generation from Diverse Viewpoints

Easi3R. [ICCV 2025] A simple training-free approach adapting DUSt3R for dynamic scenes.

slam-handbook-public-release. Release repo for our SLAM Handbook

GenFusion. [CVPR 2025] GenFusion: Closing the Loop between Reconstruction and Generation via Videos

decord. An efficient video loader for deep learning with smart shuffling that's super easy to digest

Sync-NeRF. [AAAI 2024] Official repository for "Sync-NeRF: Generalizing Dynamic NeRFs to Unsynchronized Videos"

BokehDiffusion. [SIGGRAPH Asia 2025] Official code for "Bokeh Diffusion: Defocus Blur Control in Text-to-Image Diffusion Models."

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

llama-ocr. Document to Markdown OCR library with Llama 3.2 vision

Wan2GP. A fast AI Video Generator for the GPU Poor. Supports Wan 2.1/2.2, LTX-2, Qwen Image, Hunyuan Video, LTX Video and Flux.

viber3d. A modern starter kit for 3D browser games powered by r3f and threejs -

xyne. AI-first Search & Answer Engine for work. Open-source alternative to Glean.

ever_training. Original reference implementation of "EVER: Exact Volumetric Ellipsoid Rendering for Real-time View Synthesis"

MASt3R-SLAM. [CVPR 2025] MASt3R-SLAM: Real-Time Dense SLAM with 3D Reconstruction Priors

squid-dl. a massively parallel yt-dlp-based YouTube downloader

finetrainers. Scalable and memory-optimized training of diffusion models

unilm. Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

badcad. CAD modeling for Jupyter

fillerbuster. Fillerbuster: Multi-View Scene Completion for Casual Captures

nerfiller. NeRFiller project https://ethanweber.me/nerfiller/

chatbot. A full-featured, hackable Next.js AI chatbot built by Vercel

goliath. Goliath Dataset and Official PyTorch Implementation of RelightableHands, Relightable Gaussian Codec Avatars, and Driving-Signal Aware Full-Body Avatars.

ava-256. Train universal codec avatars

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

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

mindspore. MindSpore is a new open source deep learning training/inference framework that could be used for mobile, edge and cloud scenarios.

Hunyuan3D-2. High-Resolution 3D Assets Generation with Large Scale Hunyuan3D Diffusion Models.

CE7455. Jupyter Notebook

TRELLIS. Official repo for paper "Structured 3D Latents for Scalable and Versatile 3D Generation" (CVPR'25 Spotlight).

nerfview. Python

spz. File format for 3D Gaussian splats. About 10x smaller than the PLY equivalent with virtually no perceptible loss in visual quality. Offered as open source by Niantic Labs. More details at https://scaniverse.com/spz #3dgaussiansplats #gaussiansplatting

brush. 3D Reconstruction for all

anthropic-parallel-calling. Faster and cheaper! parallel processing of Anthropic API requests, optimizing for speed, cost-efficiency, and rate limit compliance.

SplatFields. [ECCV 2024] SplatFields: Neural Gaussian Splats for Sparse 3D and 4D Reconstruction

slang-gaussian-rasterization. Slang

SAM2Point. The Most Faithful Implementation of Segment Anything (SAM) in 3D

splatt3r. Official repository for Splatt3R: Zero-shot Gaussian Splatting from Uncalibrated Image Pairs

koch-v1-1. A version 1.1 of the Alexander Koch low cost robot arm with some small changes.

kaolin. A PyTorch Library for Accelerating 3D Deep Learning Research

fisheye-calib-adapter. Fisheye-Calib-Adapter: An Easy Tool for Fisheye Camera Model Conversion

gpu.cpp. A lightweight library for portable low-level GPU computation using WebGPU.

mast3r. Grounding Image Matching in 3D with MASt3R

spconv. Spatial Sparse Convolution Library

Graphics. Source files for my experiments not limited to computer graphics.

cambrian. Cambrian-1 is a family of multimodal LLMs with a vision-centric design.

ml-4m. 4M: Massively Multimodal Masked Modeling

i-octree. [ICRA2024] Implementation of A Fast, Lightweight, and Dynamic Octree for Proximity Search

distributed-llama. Distributed LLM inference. Connect home devices into a powerful cluster to accelerate LLM inference. More devices means faster inference.

nerfstudio-webui. Gradio webui for Nerfstudio.

OpenGoPro. An open source interface specification to communicate with a GoPro camera with accompanying demos and tutorials.

xrdslam. Platform for Deep Learning based SLAM

nicer-slam. [3DV'24 Best Paper Honorable Mention] NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM

statedict2pytree. Python

PoseLib. Minimal solvers for calibrated camera pose estimation

OpenGlass. Turn any glasses into AI-powered smart glasses

lightplane. Lightplane implements a highly memory-efficient differentiable radiance field renderer, and a module for unprojecting features from images to 3D grids.

tuning_playbook. A playbook for systematically maximizing the performance of deep learning models.

wild-gaussian-splatting. DUSt3R + Gaussian Splatting

MarIOnette. Blender plugin for controlling Arduino-based microcontrollers over Serial.

imgui. Dear ImGui: Bloat-free Graphical User interface for C++ with minimal dependencies

awesome-dust3r. 🌟A curated list of DUSt3R-related papers and resources, tracking recent advancements using this geometric foundation model.

dora. DORA (Dataflow-Oriented Robotic Architecture) is middleware designed to streamline and simplify the creation of AI-based robotic applications. It offers low latency, composable, and distributed dataflow capabilities. Applications are modeled as directed graphs, also referred to as pipelines.

mvsplat. 🌊 [ECCV'24 Oral] MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images

diff-gaussian-rasterization-w-pose. Cuda

Awesome_Dynamic_SLAM. Dynamic SLAM, Life-long SLAM Research(Lidar, Visual, Sensor Fusion etc.)

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

Marigold-Video. Marigold adapted for video estimation

MLOPs. Machine Learning Toolset for Houdini

diff-gaussian-rasterization. Cuda

nanobind. nanobind: tiny and efficient C++/Python bindings

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

Awesome-Video-Datasets. Video datasets

Triton-Puzzles. Puzzles for learning Triton

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

CutLER. Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation" and "VideoCutLER: Surprisingly Simple Unsupervised Video Instance Segmentation"

jepa. PyTorch code and models for V-JEPA self-supervised learning from video.

3DGStream. [CVPR 2024 Highlight] Official repository for the paper "3DGStream: On-the-fly Training of 3D Gaussians for Efficient Streaming of Photo-Realistic Free-Viewpoint Videos".

FoVMax. Code for maximizing the total field of view for a camera array

projectaria_tools. projectaria_tools is an C++/Python open-source toolkit to interact with Project Aria data.

4DGaussians. [CVPR 2024] 4D Gaussian Splatting for Real-Time Dynamic Scene Rendering

3DGS.cpp. A cross-platform, high performance renderer for Gaussian Splatting using Vulkan Compute. Supports Windows, Linux, macOS, iOS, and visionOS

rerun. Visualize, query, and stream to train on multimodal robotics data.

pyslam. pySLAM is a hybrid Python/C++ Visual SLAM pipeline supporting monocular, stereo, and RGB-D cameras. It provides a broad set of modern local and global feature extractors, multiple loop-closure strategies, a volumetric reconstruction module, integrated depth-prediction models, and semantic segmentation capabilities for enhanced scene understanding.

Hierarchical-Localization. Visual localization made easy with hloc

resource-stream. GPU programming related news and material links

GP100L. Graphics Programming in 100 Lines

NAFNet. The state-of-the-art image restoration model without nonlinear activation functions.

inr2vec. [ICLR 2023] Deep Learning on Implicit Neural Representations of Shapes

DeepSDF. Simple and intuitive implementation of DeepSDF that you can install with a single line of code.

kubric. A data generation pipeline for creating semi-realistic synthetic multi-object videos with rich annotations such as instance segmentation masks, depth maps, and optical flow.

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

web-render-engine. 🧊 A Web Render Engine for rendering Interactive scenes with any 3D Asset from GLB, Huge-scale Point clouds, 3D Textured Meshes to 3D Gaussian Splats ✨

3dgs-avatar-release. 3DGS-Avatar: Animatable Avatars via Deformable 3D Gaussian Splatting

Awesome-LLM-3D. Awesome-LLM-3D: a curated list of Multi-modal Large Language Model in 3D world Resources

SplaTAM. SplaTAM: Splat, Track & Map 3D Gaussians for Dense RGB-D SLAM (CVPR 2024)

EfficientSAM. EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anything

EyefulTower. Official release of the Eyeful Tower dataset, a high-fidelity multi-view capture of 11 real-world scenes, from the paper “VR-NeRF High-Fidelity Virtualized Walkable Spaces” (Xu et al., SIGGRAPH Asia 2023).

supervision. We write your reusable computer vision tools. 💜

GPU-Puzzles. Solve puzzles. Learn CUDA.

SegDrawer. Simple static web-based mask drawer, supporting semantic segmentation and video segmentation with interactive Segment Anything Model 2 (SAM2).

PrimA6D. Official code of PrimA6D

Gaussian-Splatterer. Ray-traced, CUDA-accelerated Gaussian splat training utility for triangle meshes & self-contained models

Grounded-Segment-Anything. Grounded SAM: Marrying Grounding DINO with Segment Anything & Stable Diffusion & Recognize Anything - Automatically Detect , Segment and Generate Anything

self-operating-computer. A framework to enable a multimodal model to operate a computer.

nonlinear-ray-tracing. C++

supersplat. 3D Gaussian Splat Editor

sdk-examples. Spectacular AI SDK examples

dpf. Dynamic Point Fields: Towards Efficient and Scalable Dynamic Surface Representations (ICCV 2023)

recognize-anything. Open-source and strong foundation image recognition models.

Wonder3D. Single Image to 3D using Cross-Domain Diffusion for 3D Generation

CLIP-SAM. Experiment on combining CLIP with SAM to do open-vocabulary image segmentation.

BackgroundMattingV2. Real-Time High-Resolution Background Matting

aframe-gaussian-splatting. A-Frame component implementation of the 3D Gaussian splat viewer

splat. WebGL 3D Gaussian Splat Viewer

4K4D. [CVPR 2024] 4K4D: Real-Time 4D View Synthesis at 4K Resolution

ChatGPT-AutoExpert. 🚀🧠💬 Supercharged Custom Instructions for ChatGPT (non-coding) and ChatGPT Advanced Data Analysis (coding).

Dynamic3DGaussians. Python

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.

Tri-MipRF. [ICCV'23 Oral, Best Paper Finalist]Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural Radiance Fields

XMem2. A tool for efficient semi-supervised video object segmentation (great results with minimal manual labor) and a dataset for benchmarking

nerf_dataset_preprocessing_helper. Intelligent(?) image sharpness filtering for optimized NeRF training datasets

diff-gaussian-rasterization. Cuda

NeuRBF. Python

SegmentAnythingin3D. Segment Anything in 3D with NeRFs (NeurIPS 2023 & IJCV 2025)

nougat. Implementation of Nougat Neural Optical Understanding for Academic Documents

ResFields. [ICLR 2024 Spotlight ✨] ResFields: Residual Neural Fields for Spatiotemporal Signals

pyk4a. Python 3 wrapper for Azure-Kinect-Sensor-SDK

DROID-SLAM. Python

rpg_davis_simulator. Simulate a DAVIS camera from synthetic Blender scenes

NeuS2. [ICCV 2023] Official code for NeuS2

LichtFeld-Studio. Train, inspect, edit, automate, and export 3D Gaussian Splatting scenes from a single native application.

ffcv. FFCV: Fast Forward Computer Vision (and other ML workloads!)

MotionGPT. [NeurIPS 2023] MotionGPT: Human Motion as a Foreign Language, a unified motion-language generation model using LLMs

UV-Volumes. [CVPR 2023] UV Volumes for Real-time Rendering of Editable Free-view Human Performance

raytracing.github.io. Main Web Site (Online Books)

zipnerf-pytorch. Unofficial implementation of ZipNeRF

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

THuman2.0-Dataset.

DragGAN. Official Code for DragGAN (SIGGRAPH 2023)

3D-scene-and-human-capture. Combine 3D reconstruction and human motion capture .

infinigen. Infinite Photorealistic Worlds using Procedural Generation

mlc-llm. Universal LLM Deployment Engine with ML Compilation

DragGAN. Implementation of DragGAN: Interactive Point-based Manipulation on the Generative Image Manifold

nerfuser. NeRFuser

arah-release. [ECCV 2022] ARAH: Animatable Volume Rendering of Articulated Human SDFs

tava. Code for the paper "TAVA Template-free Animatable Volumetric Actors".

awesome-NeRF-papers. All NeRF-related papers @ CVPR/ICCV/ECCV/NIPS/ICML/ICLR

pandas-ai. Chat with your database or your datalake (SQL, CSV, parquet). PandasAI makes data analysis conversational using LLMs and RAG.

the-algorithm. Source code for the X Recommendation Algorithm

PetGPT. Train your own PetGPT at home!

limap. A toolbox for mapping and localization with line features.

LISST. Linear Shaped Skeleton Template for Human Motion Modelling

openscene. [CVPR'23] OpenScene: 3D Scene Understanding with Open Vocabularies

nerfshop. NeRFshop: Interactive Editing of Neural Radiance Fields

PAC-NeRF. Physics Augmented Continuum Neural Radiance Fields for Geometry-Agnostic System Identification

mega-nerf. Python

GNT. [ICLR 2023] "Is Attention All NeRF Needs?" by Mukund Varma T*, Peihao Wang* , Xuxi Chen, Tianlong Chen, Subhashini Venugopalan, Zhangyang Wang

diffusionerf. [CVPR 2023] DiffusioNeRF: Regularizing Neural Radiance Fields with Denoising Diffusion Models

XMem. [ECCV 2022] XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model

lion-pytorch. 🦁 Lion, new optimizer discovered by Google Brain using genetic algorithms that is purportedly better than Adam(w), in Pytorch

K-Planes. Python

phycv. PhyCV: The First Physics-inspired Computer Vision Library

NeRFusion. Python

petals. 🌸 Run LLMs at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading

sdfstudio. A Unified Framework for Surface Reconstruction

dycheck. Official JAX Implementation of Monocular Dynamic View Synthesis: A Reality Check (NeurIPS 2022)

deforming-nerf. Code for "Deforming Radiance Fields with Cages", ECCV 2022.

mixvoxels. ICCV2023 (Oral) MixVoxels https://arxiv.org/abs/2212.00190

VQRF. Official implementation of our CVPR 2023 paper "Compressing Volumetric Radiance Fields to 1 MB"

eg3d. Python

lovely-tensors. Tensors, for human consumption