Optical illusions using stable diffusion

Code for the paper "Visual Anagrams: Generating Multi-View Optical Illusions with Diffusion Models"

[CVPR 2025] OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities

Mobius: Text to Seamless Looping Video Generation via Latent Shift

🚀 Lightning-fast computer vision models. Fine-tune SOTA models with just a few lines of code. Ready for cloud ☁️ and edge 📱 deployment.

All-in-one training for vision models (YOLO, ViTs, RT-DETR, DINOv3): pretraining, fine-tuning, distillation.

Port of cp_kdtree to pytorch

This repo implements a Stable Diffusion model in PyTorch with all the essential components.

[NeurIPS 2024] ODGS: 3D Scene Reconstruction from Omnidirectional Images with 3D Gaussian Splatting

[NeurIPS2024] DiffPano: Scalable and Consistent Text to Panorama Generation with Spherical Epipolar-Aware Diffusion

🍳 [CVPR'25] PanSplat: 4K Panorama Synthesis with Feed-Forward Gaussian Splatting

[CVPR 2024 - Oral, Best Paper Award Candidate] Marigold: Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation

Official code for the paper "LucidDreamer: Domain-free Generation of 3D Gaussian Splatting Scenes".

Project page of "LucidDreamer: Domain-free Generation of 3D Gaussian Splatting Scenes"

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

A PyTorch implementation of NeRF (Neural Radiance Fields) that reproduces the results.

Ph.D. Thesis Template for Seoul National University (SNU)

Source code of "A structured dictionary perspective on implicit neural representations"

Official implementation of the paper "3DIAS: 3D Shape Reconstruction with Implicit Algebraic Surfaces" (ICCV 2021)

Instant neural graphics primitives: lightning fast NeRF and more

Lightning fast C++/CUDA neural network framework

Code release for NeuS

Official Pytorch implementation of "Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video", CVPR 2021

Marching cubes implementation for PyTorch environment.

Deep Multi-scale CNN for Dynamic Scene Deblurring

A PyTorch Library for Meta-learning Research